COSC 5050 Research Writing for Computer Science

1
COSC 5050Research Writing for Computer Science

• Spring 2004
• Dr. Rex Gantenbein
• Lecture 2 Comparative analysis/critical reading

2
Course information

• Reading assignment Finkelstein, chap. 2
• Title and outline due Thursday, 29 January
• See course Web page for instructions

3
Research sources

• Research starts with understanding what else has
  been done in your area
• Finding GOOD sources is important
• Background information
• Support for your arguments

4
Research sources

• Need to select sources carefully for relevance to
  both your topic AND for appropriateness as a
  reference
• Good citations help you point to information you
  dont have to include in your writing
• If citations are poor or irrelevant, they can
  undermine the validity of your research!

5
Selecting a good source

• Two kinds of sources
• Primary sources are documents or other evidence
  resulting from original research or documentation
• News reports
• Research papers (recent), technical reports
• Interviews or surveys

6
Selecting a good source

• Two kinds of sources
• Secondary sources interpret and analyze primary
  sources
• Tend to be one step removed from an event
• Textbooks
• Review articles
• Technical manuals

7
Selecting a good source

• General characteristics
• Cited in a library subject index (e.g.,
  CompuMath)
• Published by a professional organization
• Refereed or reviewed by peers (why is this
  important?)
• Includes hard results rather than unproven
  theories
• Up-to-date reference list
• Good author credentials

8
Selecting a good source

• General characteristics, continued
• Title reflects content
• Based on research or authorities in the field
  rather than merely personal opinion
• Includes supporting facts, data, diagrams, etc.
  to show that the author understands the work

9
Journals v. conference papers

• Many disciplines put more weight on articles
  published in scholarly journals than on
  conference papers
• Scientific journals tend to be archival rather
  than leading-edge
• Cycle between writing and publication is often a
  year or more
• Usually more carefully reviewed and more
  competitive

10
Journals v. conference papers

• Conference papers are usually more timely but
  less complete
• May describe work in progress, recent but
  incomplete results, student projects
• Lesser-known conferences may not review papers as
  critically, have higher acceptance rates

11
Evaluating sources

• Two basic forms of source evaluation
• Critical reading analyzing an individual source
  (e.g., a paper) for its relevance and validity
• Comparative analysis comparing different sources
  (such as journals or conference proceedings) for
  usefulness to current project

12
Critical reading

• Four factors contribute to critical reading of a
  source
• Task what the author is attempting to do (why
  the source was written)
• Subject what the author is writing about
• Audience who the author expects to read this
• Voice how the author presents the material

13
Task

• The task is often defined ahead of time (whats
  the purpose of a thesis or dissertation?)
• Present a problem to be solved
• Present results
• Describe background
• Fill in details for later use
• Other tasks may include persuasion, information,
  etc. (other kinds of writing to be discussed
  later)

14
Subject

• Understanding a subject may require additional
  reading to understand terms and assumptions
• Relates to task and audience
• Scope of subject is also important (book v.
  conference paper)
• Scope must be appropriate to the subject
• Specific gt self-contained, answer can be
  described
• Broad gt answer may never be known

15
Audience

• Consider the potential readers of a source
• Professional or lay audience?
• Conference or journal?
• Journals themselves differ IEEE Transactions v.
  IEEE Computer v. Byte
• Formal or informal?
• How do you determine the audience? READ SOME
  EXAMPLES!

16
Voice

• Choice of voice depends largely on audience,
  somewhat on subject and scope
• Personal or subjective emphasizes your own
  feelings (I felt as if ) not usually used
  for research
• Abstract or objective is more formal and detached
• More typical in research writing
• Can lead to unclear writing passive voice,
  wordiness, lack of clear focus, overuse of
  jargon/acronyms/technical terms

17
Comparative analysis

• Evaluating the usefulness of a compiled source
  (particularly a journal or conference
  proceedings) in relation to other sources
• Coherence are all articles related and relevant
  to the same purpose?
• Quality do articles come from authorities or
  respected people in the field?
• Currency are the articles addressing recent
  topics or topics of current interest?

18
Coherence

• Are all the articles addressed to the same
  discourse community?
• Are they consistent with each other (and with
  your purpose)?
• Can they be effectively contrasted if not?
• Do they contain similar levels of detail?
• Survey articles can be useful for background
  (look at their sources)
• Popular articles arent usually cited for research

19
Quality

• Who are the respected people in this field?
• Are they represented here, either as authors or
  on the editorial board?
• What is the review policy and acceptance rate?
• Usually included somewhere in the publication or
  in the instructions to authors
• Are citations reliable, relevant and accessible?
• Personal communication is not very reliable or
  accessible!
• Are Web pages a reliable source?

20
Currency

• Is the source recent, and are the articles it
  contains addressing current topics?
• Are the citations reflective or current
  knowledge?
• Older articles may be valuable for background,
  and occasionally someone may find a new use for
  an old idea
• Citations in computer science more than 5 years
  old are probably out of date

21
Characteristics of good research writing

• Everything presented here is relevant to YOUR
  research writing
• You need to understand good citations
• You need to consider your task, subject,
  audience, and voice
• You need to look for coherent, quality, and
  current outlets for your work