6

What's the best method to understand the real concept that a paper is trying to explain?

Should one go through the paper as it is presented, from first page to last page? Is there a specific walk-through to understand it faster, such as first reading the result section, followed by the concept section, and then discussion/conclusion section? Is there a best practice to be followed?

  • "How do I read?" ... What problems have you had reading papers so far? – hunter2 Jul 5 '13 at 3:16
  • Here's what I do: figures, abstract, beginning of intro, skim through results, conclusion. – cartonn Jul 5 '13 at 18:06
  • Likely duplicate: academia.stackexchange.com/q/50/73 I'll note that, as stated in the linked question, your field of research is highly relevant here, as journalistic styles differ from field to field. – eykanal Jul 5 '13 at 18:47
7

If the paper is well written, then the abstract should tell you most of the story at a high level. Then the introduction should give more technical coverage again of the whole story – you should then know the problem and the results obtained, though not all the details. The remainder of the paper then will contain just the details.

Of course, most papers do not do that, so you may also need to read the conclusion, the discussion section, and perhaps any other introductory sections. In the end, you may need to read the whole paper before you get what it is about.

However, to really understand the paper, you will need to go through it very slowly, with pen and paper at hand, and try to replicate the reasoning/ideas given in the paper, filling in the gaps – possibly with the help of additional literature.

Not all papers need to be read so deeply. Reading just the abstract and maybe the introduction and conclusion should help you decide how deep to read the paper.

  • 4
    I'd say it slightly differently: a well-written abstract should tell you if you'll want to read the paper, and a well-written introduction should help you figure out how to read the paper. But overall I agree with this. – J.R. Jul 3 '13 at 21:26
  • 1
    try to replicate the reasoning/ideas given in the paper — Better yet, try to derive the same results as the paper without looking at how the paper does it. After a good-faith effort, whether you actually succeed or not, you'll be better prepared to understand the paper's reasoning. – JeffE Jul 5 '13 at 20:00
1

NEVER READ A SCIENTIFIC PAPER COMPLETELY FROM START LINE TO FINISH LINE.

Idea is to get an understanding of what concept is presented in the paper

Open powerpoint/paper

  1. Read abstract - Tells us briefly WHAT experiment WAS DONE and WHAT WAS FOUND. What specific results are mentioned - are they relevant to your research.

  2. Discussion - Summary of important results and gives reasons based on conclusions and assumptions - Do you agree with the logic of the conclusions and are those useful to you

  3. Introduction - Motivation and importance of the research and tries to sell the paper to the maximum. Provides some background information.

  4. Results - Provides some raw data which can be related to your own research. Figures and Tables provide data in a compact format for easy understanding.

    In figures - does the graphs make sense ? , what are the axis's used and does it mean anything ? Check the units used.

  5. FINISH - I have a brief understanding of the paper and I have tried my best to cover all the relevant attributes of a scientific journal.

  6. Check the references and see if they are related to the main concept of the paper in hand. Follow through.

Manage a reference library like mendeley to keep an updated list of the literature you reviewed.

  • 1
    I would be careful with such a general claim. Depending on the field, journal format, length of the paper and skill of the authors, it is sometimes fine, in my opinion, to read a paper from start to finish. This can often be due to many of the details getting pushed into the supplemental information section. – Bitwise Jul 6 '13 at 15:56
  • 1
    Only a Sith deals with absolutes. – pratik_m Feb 3 '15 at 5:24

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.