Preface

These are some quick notes about how to read an academic paper.
The material is collected from the sources referenced here

Structure

It's good to know how a paper is structure before hand. Here we present the main parts:

The abstract summarizes

If you are sure you are going to read the paper through, it certainly is unnecessary.

It’s main goal is to present the background. Key points that is should include:

the planning of the research project
the state of knowledge in the relevant area
references to work already published
why the data and work described in the paper was necessary
clearly states the hypotheses being tested in the paper
(possibly) describes why the chosen research method is appropriate
(possibly) provides a brief description of the major answers to questions posed by the study

Description of how the research was carried out (usually compressed information).
Can be more elaborate if it’s a new/weird approach.

summary and analysis of the data
usually simple presentation and no discussion
Sometimes charts, graphs, and tables will be included here, if they make presentation of the data clearer

draw conclusions from the work described
explain what they think the data show
acknowledge limitations of the data
show how findings contribute to knowledge
correct errors of previous work
describe how the Methods and Results answer the questions set forth in the introduction

First get the gist
1. Read Abstract (if you are sure you want to read it skip)
2. Read introduction and conclusion/discussion
Read through (diagonally) the main body and figures. But don’t get stack! Try to get the flow
Go back and read it through
Go back and read again while keeping notes
Question what you read!

think about the Big question (what the whole field is about)
think about the "small" question, what they set out to answer in this paper (and where it differs from the rest)
identify the approach/method
when reading the “Method Section” try to draw a diagram
when reading the “Results”, be skeptical (do they include error bars, sample size, where they reluctant to believe it)
think if the methods and the results answer the small question
if you are interesting in the field… Read what other say about it!