[GRAD ADVICE] Reading Techniques (more) (fwd)

[AnHai Doan <anhai@cs.washington.edu>]

>From doc-talk:
-------------

        We've been publishing responses to a request from
        Janet Atkinson-Grosjean <janag@whidbey.com> regarding
        reading techniques. Here's more.
        ========================================================

From: David Atkinson <dnatkinson@ucdavis.edu>

Well I see from the responses, I am not alone feeling overwhelmed by the
task, especially since I'm not a particularly fast reader. It may take me
two days to critically read a review article.  It was two years before I
knew enough of the jargon to get up to a decent level of speed and
comprehension.  And until I had actually carried out the experimental
techniques myself, I usually just lightly skimmed materials and methods.
There was no way I could evaluate an M&M section or even understand a good
portion of it. Now that I'm near the end, I can pore over research articles
rather quickly. But the main trick in getting through a lot of literature
(and it's not easy) is finding the pony in the that vast manure pile.  I
was criticized early on by a professor for not being able to separate the
literary wheat from the chaff. But it's hard to do when you are new to a
field.  Many people I've talked to have found it useful to join a literature
club for this purpose.  Then you can get some perspective from your peers of
what's good and what's not.

By the time I got to writing my dissertation, I had fallen behind in the
literature, like so many other grad students I've known who got totally
immersed in finishing their research projects.  So I had some catching up to
do.  In desparation I came up with an approach that works pretty well for
me.  For each major topic, I found four or five review articles and outlined
the relevant (to me) parts of each one with a word processor, guided by the
headings in the articles.  For each important point or discovery, I put the
corresponding citation(s) next to it in the outline. Then I colored each
outline using the character formatting feature of the word processor. Next I
merged all the outlines, combining related points and kept a color key at
the bottom so I could refer back to the original review article.  This made
it easy to see which original publications were frequently cited as well as
those that seemed to make a significant contribution to the review (often an
alternative explanation or contradictory experimental result).  It also gave
me a historical perspective with a short reading list, allowed me to
see who were the big guns in the field and provided a framework for
following the field in recent literature.  It took me a day or two to do
each review, but the result is a very handy quick reference guide for each
major topic.

Dave Atkinson - "Almost done...just a couple more weeks, really"

==============

From: "Jan Spears" <JSPEARS@wtamu-math1.wtamu.edu>

I use the same method as Alex Vrenios (read the abstract and the conclusion). 
If the material is relevant, I then skim the article noting the main points
and the overall experimental results. I record that and the bibliographic
information in a HyperCard stack.  I have one card for each article and I
can use the stack call up lists of articles by author, topic, or date.  

For those articles which my profs have identified as "significant in the field",
I reread them periodically in the hopes that by the time I get to my defense, I 
will be _very_ well-versed in their content and meaning!

Dept. of Mathematics, Physical Sciences 
         and Engineering Technology
-----------------------------------------------------------------------------