We are in the right direction. As a researcher/ masters student/ working professional we have to read research papers. We have to keep ourselves updated based on the technological advancements happening around us. Believe me, we are living in a highly competitive world and this is more important than you think.
That being said we can jump into today’s blog. Everyone knows how to read, the interesting fact is how effectively we can read it. Once I was having a discussion with my mentor and he said in their lab people were practising reading speed which is reading words per minute without having a lot of eyeball movement. Does not that sound crazy 😱, however, it is true. Albeit we are not trying to achieve that level, let us try to understand what all thing we should pay attention to while reading a research paper.
How to read a research paper by S Keshav from the University of Waterloo is what we are going to look into. Learning to efficiently read a research paper is a critical but rarely taught skill. The three pass approach will help us to estimate how much time need to allocate and the depth of the paper. This will also help us to get a bird’s-eye view. Rather than getting lost in the plethora of information in the research paper.
THE THREE PASS APPROACH
Pass one
- Carefully read title abstract and Introduction
- Read the section and subsection heading but ignore everything
- A glance at the mathematical content if any to determine the underlying theoretical foundations
- Read the conclusions
- Gance over the references, mentally ticking off ones you have already read
The idea is at the end of the first pass you should be able to answer the following
- Category: What kind of paper is this? A measurement paper? An analysis of an existing system? A description of a research prototype
- Context: Which other paper is it related to? Which theoretical bases were used to analyze the problem?
- Correctness: Do the assumptions appear to be valid?
- Contribution: What is the paper’s main contribution?
- Clarity: Is the paper is well written?
Once you have done all of these you will get an idea about what to do next with this paper. Either you can discard it or give more attention and go to the second pass. Also, this idea can be kept in mind while you write a paper as well. You can think as your readers will only spend 5 minutes on reading these things and if they can’t perceive what is the gist of this paper then it is a failure adding a diagrammatic description that has more detailed expression would also support the readability of the paper
Pass two
Here we are reading the paper with greater care, you can take notes or highlights points that seem important to you. Also, make sure that you note down all the terms that you don’t understand or the questions that you want to ask the author.
- Look carefully at the figures, diagrams and other illustrations in the paper. Pay special attention to graphs. Are the axes properly labelled are results shown with error bars. So that conclusions are statistically significant? Common mistakes like these will separate rushed, shoddy work from excellent
- Remember to mark relevant unread references for further reading ( improve your background)
The second pass should take up to an hour for an experienced reader. After this, you should be able to understand what exactly the paper is talking about (the content). You should be able to summarise the main thrust of the paper with supporting evidence. Please keep in mind that this level of ponderation will be needed if the paper is in your area of interest. If it is not, you might not understand what exactly in this paper even after your second pass.
In that cases according to the author, you can do one of the following:
- Set it aside hoping that you don’t need to understand this paper in this detail
- Return to the paper after reading the background
- Persevere and go on to the third pass
Pass Three
You need to understand the paper thoroughly, especially when you are a reviewer. The easiest way to understand a paper better is to implement it or recreating it and then also do a comparison of the results that you got to that presented in the paper. It will help you to find the hidden failing and assumptions as well
This part needs great attention to detail. You have to note down each assumption and statement. And the main idea will be how you will be presenting the same idea if you were the one doing the project. Also what all will be the future work as of your thoughts.
This can take many hours for a beginner like us and 2-3 hours for an experienced reader. Once you are done with this pass. You will be able to reconstruct this paper from your memory. And also the things that missed by the authors in this paper like citations and assumptions
So now we can answer a big question How to do a literature survey?
This is where we can put what we learned in this paper to practice. First, go to Google scholar or Citeseer and select the good keywords to find the paper. From the results select highly cited papers in the area
Do pass one and understand the sense of their work. Then read their related work session as well. If you can find a survey it is the best. Otherwise, find the shared citations and authors which will navigate you to the current researches in that area. Then go to the top conferences and look through their proceedings and the high-quality work and the work from the citations will give you a survey. Make two passes through them, and if there is any key paper that is being cited in your survey find and read them as well and to the iteration as necessary.
And as we grow we might need to do reviews of the paper as well for that we need to read another paper by Timothy Roscoe’s paper on “Writing a review for systems conference”. 🔖
For technical papers Henningn Schulzrinne's comprehensive website and George Whiteshides’s excellent overview of the process. Also, Simon Peyton Jones has a website that covers the entire spectrum of research skills. 🔖