1

I'm writing my masters thesis at the moment. The topic of my thesis is the evaluation of one classification algorithm. Due to its technical nature I have to give a lot of context for each experimental setup. Something like the choice of signals to be classified for example:

The signals tested in this experiment were fsk signals with the following technical parameters:

Shift/Symbolrate

  1. 500/75
  2. 500/150
  3. 500/200
  4. 500/250
  5. 350/150
  6. 350/75 Each of the signals contained 20 header information and 80% payload

500/75

Due to the thesis containing about 20 experiments of various sorts, I'm not sure on how to best present my information concisely.

Would you try and make a subsection for each experiment, use a table or present it in plain text?

  • 1
    While this is certainly on topic here, I think you might get more informed answers on the Academic SE. But for a thesis, the point of which is to allow your examiners to assess the quality of your knowledge and thought, being concise in the presentation of data seems like a pretty low priority. I would be inclined to focus on making it as easy to examine as possible. – user16226 Apr 6 '17 at 12:05
1

The usual approach is a set of tables. This is a paper, you don't try to be interesting, you want to be informative.

If you don't want the "main gist" to be lost among plethora of tables, you can perform a top-down "decomposition", where after introduction of methodology you present first a brief summary of results (which you then repeat in the "Conclusion" section at the end), and then follow up with how you arrived at these, through tables of various experiments, one chapter/section per experiment, as many tables as relevant to the experiment included.

Regardless of all, tables are the usual flesh of any scientific papers on experiments, presenting their results. Methodology, background, theory, that's all "fluff" that helps people in understanding the results, proves their validity and roots them in the "general image", but they read the paper for these results, and these will appear either as tables, or as graphs, which, while more evocative, are less precise (so you may include both.)

Look up some white papers on material engineering. A typical approach is two-three pages of "background", followed by 50 pages of tables.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.