3 ms·
Ask HN: How do you read a research paper?
Whenever I open an interesting title on a research paper, I get bored quickly because there is just so much to read and so many different things I don't understand like the maths behind it. But the topic is still interesting, I just don't get how I should read such paper.
How do you do it? Do you actually read the whole thing?
What's the technique to extract the juicy content of it? How should I think and treat the paper?
- Quinzel 3y agoProbably if you haven’t really learned how to read the result section, then it’s better to read the abstract, introduction and then skip to the discussion/conclusion. If you don’t really understand the method or results though, it’s harder to know if what you’re reading is pure bullshit. I actually had to be taught how to read and understand the results sections of research papers and I would rate my current skills at like a real basic level, and only in psychology where I have a basic understanding of the descriptive statistics, effect sizes, correlations, regressions and a basic understanding of factor analysis. I’m still actually working on having a deep enough level of understanding that I could critically analyse a research paper. There’s other ways to ascertain whether a paper is garbage, like if it was published in a journal with a low impact factor and a high acceptance rate for publication - that can be an indication, but really unless you can understand the lingo and the sometimes very mathematical results section it can be quite hard to know if what your reading is worth the time of day.
- syndicatedjelly 3y agoRead the abstract, intro, discussion and conclusion first. If anything stands out to you, find the methodology and read that. Save the results for last, only read them if you find yourself disagreeing with any conclusions the author came to
- brudgers 3y agoReading academic content is a skill that is learned through practice. And learning is a skill that is acquired by being really bad at something but doing it anyway. And accepting that there is a lot of ground to make-up just to get a basic understanding...like all the math you mention. So you learn more math to better understand things with math (and of course that means being bad at the math you want to learn). On the positive side, there's no test next week and in a few years you might better understand the thing you are struggling with today. That's what growth is and sure growth takes time. But time is all we have. Good luck.
- gus_massa 3y agoI usually read the abstract and then look at the tables and graphics. If they look interesting and convincing I read the rest. Why do you want to read the paper? Are you panning to implement it? Just to learn general information? Make angry post in the internet?