Checklist Scientific Paper

From its-wiki.no
Jump to: navigation, search

Checklist Scientific Paper

Before writing, create a table of content (TOC)

  • define a red line: what is the major story of your paper, what do you want to demonstrate/show?
  • split that into sections
  • put under each section/subsection a list of bullets on what to write
  • Recommendation: use comment option in LaTeX to indicate the content, e.g. % analysis of first example, use only metrics for a and b

Define on forehand a semantic, and be consistent in your wording

  • approach or a methodology
  • analysis or an evaluation

Define a clear baseline/starting point of your approach

  • what do we have today?, e.g. multi-metrics approach
  • what are the deficiencies? - e.g. we are missing a complete system description
  • comparison state-of-the-art in research
  • what am I supposed to add? - e.g. complete system description. still some simplifications

Specifics

  • check each sentence:
    • does the verb reflect the right relation to the object, e.g.
    • simple constructions, use only 1 object with max 1 attribute, e.g. contributes to an extended approach
    • avoid this. Wherever you use this, reflect on whether it is "crystal clear" or leaves doubt. Use rather: The approach ...
    • tables, figures need to have different captions.
    • every time when you use «this», make sure that you really refer to one thing. Otherwise, repeat rather the subject than use «this».
    • make yourself a list of relations, e.g. «evaluation of two components are equal» %incorrect: result of an evaluation of two components lead to the same ...
    • check that you reduce the number of objects in a sentence. a verb relates to one object, having in maximum one adjective.
  e.g. system components interconnection graph -> an interconnection graph for system components


Evaluations

  • evaluate your approach with respect to e.g.
    • sensitivity (what changes if I change an input value)
    • cost/benefit analysis - there is no such thing as a free lunch
    • after the analysis, are there lessons learned, things you would have made differently?

Conclusions

  • start with a very short summary of what was performed
  • use numbers, be specific
  • pick up on what you wanted to show/demonstrate
  • conclusion is a : a) summary, b) facts


References