On time and reading articles
A general rule of thumb at the university is that an average course
workload is 2 hours outside class for each hour inside (for us that
would be 6 hours a week outside class). For students with a 5 course
load, this works out to 45 hours a week (9 hours per course, 3 in class
6 out). Many courses have large projects so instead of looking at
the 6 hour outside per week, you end up looking at the total of 13*6
(108) hours outside per semester. However, CS4472 is designed toward
a relatively steady workload rather than a rush at the end of the
In particular, an assignment like `read a 10 page paper and answer
a reaction question about it' is viewed as a 2 hour outside class task.
Of course, one could spend a semester just on trying to get the most
out of the one article (and investigating how it connects to the
articles it cites and those that cite it) as well as trying to figure
out the best answer to the reaction question and how to most wisely
use the 100 words to get across the best answer, but I would instead like
you to think about spending 2 hours reading the paper and trying
to figure it out and answering the reaction paper and bringing questions
and difficulties that are not yet resolved to class to be sorted out as
best we can during class time. While two 2 hour tasks still leave 2
hours unclaimed each week; between two 1-page proposals for 5 hour projects,
doing two five hour projects, writing up two five hour projects, and
studying for four quizzes, those extra 2 hours (26 hours across the
semester) will probably get used up.
So, the question is how do we end up with 2 hours to read a 10 page
paper like the category partition paper of last week. Well, first
we consider how long it would take to read it out loud. I timed page 678
of the article (a page with no figures, so the longest to read) and it
took just under 8 minutes to read it out loud (reading out loud makes
sure I don't skip over anything -- much of the material can be read faster
than `out loud' speed although some of the more technical aspects are worth
pondering a bit). At 8 minutes a page, that is 80 minutes or 1 hour and 20
minutes to read the paper. To get the most out of your reading, I would
suggest the following
- Read the reaction question first, so you can make note of parts of
the reading that directly address on the reaction question.
- What is the author claiming to do? Does the author actually do it?
Read title and abstract to get an idea on the first question and then skip
to the summary/ conclusions section at the end of the paper to get an idea
on the second question. The classical approach to writing is to tell people
what you are going to tell them, tell them, and then tell them what you
told them. Usually the last phase does a better summary than the first
phase. Similarly, the central claim of the author is generally known as
the `thesis statement' and often appears as a sentence in both the summary
and main body of the presentation, but sometimes is missing from
poorly written abstracts. Knowing where the author is going generally
makes it easier to follow how the author gets there.
- Can you make use of this? In general, this is a very
applied course where what we are talking about should be of use to
a practicing programmer. But more specifically, the presentation
of material has been focussed on introducing techniques that you
might use in a project in the early part of the semester, so these
papers are intended to tell you stuff that you could actually make
- Pay attention to technical vocabulary The concepts
that an author thinks are particularly important usually end up
being given a name. New vocabularly is often italicized in
these articles, but sometimes not.
- Pay attention to examples Again, authors don't usually
give an example unless they think it is important to get a point across.
- Pay attention to topic sentences Just as most articles
have a central thesis statement, each paragraph usually has a central statement
(the topic sentence) around which the rest of the paragraph expands.
Usually the topic sentence is the first sentence of a paragraph.
Thus, a quick overview of the paper can often be gotten by just reading
the first sentence of each paragraph.
- How does this paper fit in with the other readings?
- Again, vocabulary can be an issue, both when it is shared between
papers to indicate a linkage, and when two authors use the same terms
differently. For example, both `environment' and `functional testing'
are important terms in the first two papers. Do both authors mean the
same thing when they use these terms?
- If we were going deeper into the topic, the Bibliography and
sections on Related Work would be of interest, but for the purposes
of this course, related work is only of interest if the related work
was part of the assigned readings. Since we have read the Whittaker
paper and it directly references the Category Partition paper, it
is worthwhile looking into just what Whittaker had to say about
this paper. Even when there is not a direct citation, it is sometimes
helpful to note references to the same author (or points where the author
references their self). Keep in mind that not all authors believe the
same things and so keeping track of related authors is one way of keeping
track of discussions based on consistant assumptions.