Thursday, June 26, 2008

Your Home From Home

I got back from a conference in London this week, where my shabby little
dorm room let me try out a principle I thought up the last time I was
staying away from home. At that time I noticed my bed & breakfast room was
a horrible mess, after only 2 days, and I was having trouble finding
stuff. I realized there's only one solution to that for me - the same
exact solution as for me at home. And that's attention to flow, and
processes. I realized I should set up a mini version of my home system
whereever I am, *even if only for one night*. I already know how to use
it, and so it's just a matter of setting up each of the components. And
here is that minimal set:

* A place for clean clothes
* A place fpr dirty clothes
* A clothes chair for the ones that don't fit either category (see the
limbo laundry entry)
* An inbox (for when I come home from a day at the conference and want to
just dump stuff - in this case it was just the corner of my desk)
* A garbage and/or recycling within easy reach.
* An out box - a location I put stuff that I want to remember to take the
next day.
* A "going home" out box. This is the only element that is special to the
away from home scenario. Here I put stuff I don't need to think about till
it's time to go home.
* A "temporary support" location, for things that I will need to make use
of for short-term projects.

These show that I'm now definitely leaning towards the "take everything
out of your suitcase right away" side of the issue. In about 15 minutes,
once all those things were in place, my UWO dorm apartment was a
functioning helpful organism, just like my home and office (well they are
that some of the time)

One additional thought I just had: if you're remembering various things
you want to pack a few days before a trip, why not toss them in your open
suitcase? It's taking up space anyway, might as well use it as a bin -
then the distance to move the stuff before you repack will be particularly
short.

Monday, June 09, 2008

Colour Coding to Help Intense Reading

A great contribution from friend and fellow grad student Liz Arsenault:
--
Our lab holds small journal club meetings once a week where one member
leads a discussion on an article of interest. Everyone reads the article,
but only the presenter has any obligation to be very familiar with it.
When I am preparing to present an article, I like to underline the points
of discussion in colour-coded felt-tip pen and highlighter.

The procedure is way too intensive and time consuming for articles that
just need to be skimmed, but I've found it very helpful when I need to
know an article inside and out.

1) I colour code premises, methods, results, people, terms, and
definitions*. This allows me to quickly locate the answers to things that
can slow down discussion like misunderstandings on the finer points of the
methods, clarify terms/acronyms, etc.

2) Colourful felt tip pens are something I find motivating. It's fun to
do, and it's fun to look at.

3) It makes the amount of work you've done obvious to your supervisor ;).
Anyone can underline anything that looks like a declarative statement in
pencil. When it's in a lot of colours, it's clear you've done your
homework.

4) Reading looking for these sorts of information makes me approach an
article in an appropriately analytical frame of mind. Unfortunately, I
can't read science as easily as a novel.

5) Tracking people (authors, big citations) makes it possible to build an
understanding of which people work together championing which sorts of
ideas, and reminds me to look at author webpages to look for other papers
I might find relevant.

6) I'm trying to work out how to go through all my term and definition
highlighting to build a personal glossary of my area of study, and this
should also help me to detect when people use same terms to indicate
different things or when they use different terms to indicate the same
things (the latter happens a LOT between psychology vision papers and
computer vision papers).**

Granted, there are major headings for results, methods, conclusions in
most articles, but often some of these things are phrased most succinctly
in the abstract or results in the discussion, etc.. Plus, when you're
flipping through several pages (as long as you aren't colour blind),
looking for pink pen is a lot quicker than skimming looking for the
"methods" heading.

I will let you know of my other brilliant processes as they develop ;).

Take care, Liz

* Order of importance:
1) terms
2) premises
3) results
4) methods
5) people
6) definitions
... the lower priority elements get dropped depending on how quickly I
want to process the article.

** A psychological benefit of doing this is seeing yourself highlight
fewer and fewer terms as you become more proficient in an area of study.

Friday, June 06, 2008

How Obama Used Feedback

From a fascinating New York Times profile a few days ago:
(one question: what does "favoring a hand on the shoulder" mean?)

--

"You don't go from being a community organizer to running for president in
15 years unless you have a lot of ambition," said Paula Wolff, a Chicago
Republican and a mentor. "He likes to listen carefully, and naturally you
assume that's very smart of him."

If there is an art to seeking advice, Mr. Obama holds a master's degree.
He favors a hand on the shoulder, a whisper in the ear. In 1996, when he
pondered a race for the Illinois Legislature, Jean Rudd, a mentor in the
foundation world, took him to lunch with a prominent lobbyist. The
appetizers had no sooner arrived than the lobbyist framed the question:
Why would a Harvard-educated lawyer want to step into a hellhole like
that? You'll leave your wife behind, you.ll be in the minority party,
you.ll be treated like dirt. Mr. Obama chuckled and asked questions. The
lobbyist later became an adviser.

Abner J. Mikva, the former judge, asked Mr. Obama, fresh out of Harvard,
to apply as his clerk. Mr. Obama declined, preferring to labor as a
community organizer. But, characteristically, he later befriended the
older man.

The judge recognized his talents, but oh that speaking style. Too many ers
and uhs, too Harvard and not enough South Side. Mr. Obama did not argue
the point; he began paying attention in church. [to learn public speaking]

--

By contrast, in a jawdropping interview on Fresh Air Scott McLelland
revealed that Bush hates to have opposing viewpoints from his own
represented in his advisory groups, and never reads opinion pieces he
doesn't agree with.

Monday, June 02, 2008

Life Advice from the Last Lecture

"Carnegie Mellon Professor Randy Pausch, who is dying from pancreatic
cancer, gave his last lecture at the university Sept. 18, 2007, before a
packed McConomy Auditorium. In his moving talk, 'Really Achieving Your
Childhood Dreams,' Pausch talked about his lessons learned and gave advice
to students on how to achieve their own career and personal goals."

http://www.youtube.com/watch?v=ji5_MqicxSo
(it's about 1 hour and 15 minutes, and enormously engaging and moving)

Much more than the time management lecture, these notes are no substitute
for watching him. Partly that's because what might seem like flat
aphorisms are just the punchline to rich stories from his career. And
partly because what really strikes me is the metadata: the context, the
manner of speaking, the life itself. My reaction was, this is a nerd made
good. Unlike some of us analytically-minded guys who are attracted to
computer science like he was, he reached outside himself, to find what are
the really important things that make up a life. And over and over he
defines it in terms of human relationships - technology is a secondary
(though important) player. There are three major things he shows off, that
I am currently trying really hard to incorporate into myself, and wish I
had started long ago: Appreciating people, being positive, and defining
and expressing who you are so strongly. Right at this moment, would I be
able to express so strongly who I am and what I'm about, what my story is,
if I suddenly had my own opportunity to give a Last Lecture? What an
inspiration.

---

Brick walls are there for a reason: they let us prove how badly we want
things. They're there to stop the *other* people

Fundamentals, fundamentals, fundamentals. You've got to get the
fundamentals down, or else the fancy stuff won't work.

When you screw up and no one's bothering to correct you anymore, that
means they've given up.

When you do something young enough and train for it, it just becomes a
part of you.

Experience is what you get when you didn't get what you wanted.

It's very important to know when you're in a pissing match. And it's very
important to get out as quickly as possible.

That's one of the reasons you should all become professors: It's because
you can have your cake and eat it too.

Course called building virtual worlds: 50 students from all different
departments. (art, design, drama, and CS) Randomly chosen 4 person team,
change every project ("three new playmates"), two week projects, 5 per
semester.

What to do when they completely blow you away on the first project? "That
was pretty good, but I know you can do better." You don't really know
where the bar is, and you're doing them a disservice by placing it
anywhere.

ETC curriculum: "5 small projects followed by 3 big projects. All of your
time is spent in small teams, makin' stuff."

Project-based curriculum
Intense, fun student experience
Field trips!

Alice: "Millions of kids having fun while learning something hard. And
that's pretty cool. I can deal with that as a legacy."

He painted stuff all over his bedroom walls. To anybody out there who's a
parent, if your kids want to paint their bedroom, let them do it.

"Randy, it's such a shame that people perceive you as so arrogant. Because
it's going to limit what you're going to be able to accomplish in life."
What a hell of a good way to word "you're being a jerk."

Particularly with middle school girls, if you present it as a storytelling
activity they're perfectly willing to learn to program computer software.

Decide if you're tigger or eeyore
Never lose childlike wonder
Help others

How to get people to help you
- Tell the truth
- Be earnest (I'll take an earnest person over a hip person any day,
because hip is short term but earnest is long term)
- Apologize when you screw up
- Focus on others, not yourself

Get a feedback loop, and listen to it. When people give you feedback,
cherish it and use it.
Show gratitude.
Don't complain, just work harder
Be good at something, it makes you valuable
Find the best in everybody. It might take years, but people will show you
their good side.

[the lecture is ] not about how to achieve your dreams - it's about how to
lead your life.


Wonka :: "Do you know what happened to the boy who got everything he ever
wanted?"
Charlie :: "No, what?"
Wonka :: "He lived happily ever after."

--

Finally, from his equally moving (and much shorter) commencement address
at Carnegie Mellon this May:

You will need to find your passion. Many of you have done it, many of you
will later, many of you will take until your 30s and 40s. But don't give
up on finding it. Because then all you're doing is waiting for the reaper.
Find your passion, and follow it, and if there's anything I've learned in
this life, it's that *you will not find your passion in things.* And you
will not find that passion in money. Because the more things and the more
money you have the more you will just look around and use that as the
metric, and there will always be someone with more. So your passion must
come from the things that fuel you from the inside... that passion will be
grounded in people. It will be grounded in relationships with people, and
in what they think of you when your time comes.

My Citation Management System: Refinements

Unlike the system I talk about in the 3 previous blog postings on citation
management, these are things that are out there on the fringes of my
system, stuff I'm still tinkering with.

--

* My someday/maybe reading list was getting too long and intimidating, so
I started putting in a fake entry at the end of every month to divide it
into manageable monthly chunks. So may just passed, I created a new blank
reference in there with the title "2008 MAY" so all references I added
between the start of may (designated by "2008 APRIL") are nicely
delineated.

* My physical paper management system, which predates my general filing
system, was breaking down. I keep papers in hanging files by topic area.
If I were starting all over again I might use Jim's system of putting them
in manila folders by first author name in alphabetical order, but it's
nice to be able to scoop up a whole handful of a particular type of paper,
and hanging folders are much more sturdy. That's the problem though,
some are bulging. (though like with my manila folders, I make sure to have
plenty of empty ones within arms reach, so I can create a new one on the
spot) A couple of embarassing incidents of not being able to find a paper
made it clear I need to do better in indexing my printed-out papers.

I created a new field called Filed under, where I enter in the heading on
the tab for the hanging file it's in. What really makes this work, makes
it quite practical to (eventually) go through all my hanging files and
enter the Filed under in the corresponding citation entry in my Read
database, is a neat feature of endnote called Term lists. With just a
few minutes poking around in help and menu items, I figured out how to
make it so I would only have to type the first letter of a hanging file
name, and the rest of it would pop up in the field, just like it works for
author names in endnote.

So now whenever I have a few minutes when I have absolutely nothing else I
could be doing, I grab a couple of my hanging files and make sure all the
papers have entries in my databases and have their Filed unders filled in.
I also sort the papers in each hanging file by first author. So now (for
the ones I've done anyway), it's a snap to go from the citation entry to
the physical paper.

* Sometimes in doing a lit search I run across papers that are completely
irrelevant to my research but have too good titles to let slip away
forever: "A triangular theory of love" "Does Aerobic Exercise Decrease
Brain Activation?" I can stop myself from reading them on the spot by
throwing them into a new database I made called Interesting.

* By painful experience, I can easily read a paper and then have
absolutely no recollection of it just a couple of months later. Therefore
though it's cumbersome, I *must* keep notes on the papers I read, at least
if I'm ever going to go back to them. I tried many different ways to do
it, some electronic, some paper, but the best thing so far involves using
endnote again. In the Research Notes field, I type in what I call the
Gist. This is as few words as I can type to encompass everything I would
want to refer to this paper for. Sometimes that's a lot, including lots of
experiment details. Sometimes it's one sentence. However it almost never
looks anything like the abstract: first of all its in my own language
("...therefore, they say, feature integration theory is whack.") and
second it's what I got from the paper that was new and semi-surprising,
not everything that was in there. So it's customized to me and my
interests.

This takes effort for each paper that I read (though it doesn't mean I
have to read at the computer: I tend to jot brief notes at the bottom of
pages with a pen that I can then mindlessly transcribe into the system
later) but ultimately it saves a great deal of tiresome effort. Recently I
had to review an area (rapid visual categorization) and I was almost shed
tears of gratitude realizing I had filled in the gists for all 3 of the
key papers, and so would not have to even glance at the originals, let
alone read them through. A job that could have taken 2 hours (assuming
total amnesia, which is far from impossible) instead took 15 minutes.

Of course this means my endnote Read database is even more precious and
needing of being backed up.

* Somethign that goes great with entering gists in EndNote is creating
custom output styles to help generate neat reports on groups of papers.
It's easy to make a bibliographic list in APA format, just by selecting
multiple citations and choosing Copy formatted. But now imagine the power
of being able to create an output file with all those citations *plus
their abstracts*. This is easy to do: just make a copy of say APA 5th
style, go to Output styles -> open reference manager (this is all EndNote
X specific), click to edit your style, go to Layout under Bibliography,
and in "End each reference with" click Insert Field and add the Abstract.

I created another one based on APA 5th called Gist Output, which lets me
see the gist info that I've entered. It's particularly great for seeing
output in the preview pane when I select a reference in the window. Rather
than using the Layout mechanism, I changed the Bibliography templates
directly. So it looks like (this is for the Journal Article entry)

Author. (Year). Title. Journal.

Research Notes


(with title in boldface, and journal in italics) The disadvantage is that
you have to repeat those changes for all the different reference types you
might want to use.

Hacking output styles is great, because it lets you quickly and easily
output sheets of info, including gists/abstracts, for any subset of papers
you want. You could print out pages with all the abstracts from Trends in
Cognitive Science for a year, a great way to skim the literature. You
could choose two or 3 to have in gist form for an important meeting. Etc.