How to read the literature efficiently?
At the beginning of doing scientific research, perhaps everyone will be frightened by the three words “read the paper.” If you don’t read the paper properly, it is estimated that the weekly group meeting will be tormented, and the discussion with the tutor will be difficult to communicate.
This kind of experience sounds very persuasive, but there are methods to follow in paper reading. After doing scientific research
as a student and tutor for a period of time, I deeply felt the importance of the paper. At the beginning of scientific research, it is necessary for every
student to read some papers intensively, understand the whole research in mind, as if he had done it himself, and then enter the laboratory without panic. Why can’t
I read the paper? If things go on
like this, many people will inevitably come up with the idea: Why do I have to read these papers? Why can’t
I read the paper?
How to find the right paper to read and how to read a good paper are the core of “reading a good paper”.
But the premise of achieving both is to lay a good foundation for scientific research.
Without enough scientific research foundation, you can’t understand the knowledge points of the paper, and you can’t control which papers you should spend more time reading.
In fact, the paper can not see, mostly because of the following three reasons: fear of difficulties (psychological factors) at the beginning of reading the literature will have fear of difficulties, often manifested as: fear of reading: the thought of reading papers on anxiety, and even some students back to the generalization of fear of learning places; Procrastinate reading: When several things appear at the same time, they will do other things to avoid reading the literature; only choose simple reading: Although they start reading, they prefer to find simple literature to read when they are faced with multiple papers. The main cause of
these psychology is that after entering scientific research, they are neither fully prepared nor adapted to such high-intensity work.
Apart from the fact that some students are afraid of learning from an early age, another reason is that they overestimate the difficulty of scientific research.
In fact, scientific research is also the completion of a project.
To overcome difficulties, we can resolve insurmountable problems into a number of small goals that can be easily accomplished, and dismantle every detail of scientific research, as long as we can do it step by step. Lack of
English ability (skill factors) to make up for the lack of English ability can not be achieved overnight, need to accumulate step by step, step by step can improve reading ability. To
improve your English ability, you should start with the “English skills needed for reading literature.” If you can pass CET-4 and CET-6, you will have already mastered the basic knowledge of English grammar.
However, English literature is much longer than reading comprehension, so you should be prepared to improve your English ability for a long time. The problem of
not understanding why to read (target factors) is the same as “what is scientific research.” This kind of psychology is actually caused by unclear goals and mismatches.
Many students have the thinking inertia of “I’m working for the boss” and “I don’t like this scientific research project”, which leads to the resistance to literature reading, which is also the most essential reason why most students feel that their papers “can’t go on”. Why
Scientific Research Should Start with “Reading Papers” Doing scientific research in college is very different from doing coursework in junior high school. In fact,
high school and junior high school assignments all have “standard answers.” It can also be said that all the knowledge points are in books, and can be completed as long as they are done step by step.
This characteristic will negatively affect some students’study in the university stage, because the scientific research in the university is directly at the forefront of scientific issues, that is, all the content that is not in the textbooks, even the professors do not know, otherwise where is the exploration of the unknown?
Coursework doesn’t create something out of nothing, it’s just repeated practice. And university research requires “digging three feet into the ground,” requiring you to discover things that your predecessors have not discovered, or even completely unexpected.
Literature is the “shoulder of a giant.” Only after reading the literature fully can one understand the context and progress of scientific research, and finally begin to experiment and explore. The amount of papers you have in
your heart (remember, “printed” is not the same as “read”) directly determines the breadth and depth of your thinking about scientific research issues, so the ability to read papers can be said to be the basic skill of scientific research.
Take my own field as an example, a good journal paper should have about 40-50 basic citations.
But that doesn’t mean you only need to read 40-50 papers. In order to complete the background knowledge of the subject, you actually need to read 80 papers.
If you want to do your own research, you have to read more.
These papers are not necessarily related to your direct research, but they can lay the foundation for your knowledge system.
How to read a good paper, how to find a suitable paper to read, and how to read a good paper are the core of “reading a good paper”.
But the premise of achieving both is to lay a good foundation for scientific research.
Without enough scientific research foundation, you can’t understand the knowledge points of the paper, and you can’t control which papers you should spend more time reading.
How to find the right paper First of all, just enter the laboratory, especially when the tutor gives the direction but does not give the specific paper, do not immediately go to find the paper.
When you have just started scientific research, you do not necessarily have the taste to judge excellent papers. If you read too many bad papers, won’t the whole foundation be distorted? When I
started my master’s degree, I didn’t know what kind of paper to read, so I searched and downloaded some literature related to the content mentioned by my tutor directly in the academic search engine and began to read it. Looking back
after a long time, these papers can’t be said to be “bad,” but they are also not in good taste. The journals they publish are not good, and the paradigms and problems they study can’t be said to be particularly valuable.
Therefore, many papers found blindly in the early stage of scientific research are of limited value, which may not only spoil your research taste, but also waste a lot of time, and not learn much.
From my own example, if you want to enter the academic state efficiently, you might as well start from three directions: (1) Understand the context and taste of academic development from classical textbooks. When I was doing scientific research seriously, I first read a scientific research book by my tutor’s tutor, which contained all the achievements of his 40 years of scientific research. It not only combs the development of the field, but also sees the inheritance of teachers, and contains a lot of first-hand literature information. After
reading it, I can not only quickly know the existence of these documents, but also understand their location on the academic knowledge network, and through his comments, I can understand “what is a good paper and what is a paper with excellent taste.”.
(2) Starting from the papers of our own division, we can understand the methods and research fields of our own laboratory. There are only a few clear lines of inheritance in our division.
So why not take advantage of it?
At that time, I quickly brushed through more than 100 papers in my own laboratory (that was in 2013, 257 so far).
Fortunately, my tutor likes website construction very much. All her papers are filed on the official website of the laboratory. Therefore, I can quickly consult all the papers and find the right ones. After reading them, I can communicate with my tutor immediately and get feedback directly. Learn literature
from fellow students. However, tutors are bound to be busy, and sometimes they are too embarrassed to bother them directly with questions that are too simple.
Therefore, when there is no direction to find the literature, it is a good way to directly ask the same door which expert’s paper should be looked for. At that time, there were two doctoral brothers and sisters in
my laboratory, as well as several postdoctoral sisters, and the environment was particularly good.
I still remember that one afternoon several years ago, my senior sister told me to read the papers of Professor Todorov of Princeton University, and explained several of his representative works to me, which changed my research direction to a certain extent and led me to the research direction of this small field. What
if you are going to dabble in an area that your lab is not familiar with?
For example, many cognitive psychology laboratories have begun to explore the integration of machine learning algorithms such as DCNN and human cognitive processing systems. Obviously, the tutors and teachers in these laboratories are not familiar with these fields with a large span. It seems that neither the same school nor the tutor can seek help. Is it necessary to post online for help at this time?
Not really.
First, go back to the textbook. Textbooks must exist for
mature and basic subjects.
Here’s a way to find papers: classic papers contain classic paradigms.
Although these papers are indeed far from the front line, any search on any academic engine can find out who has cited these papers recently.
After that, you can find papers that cite classics in recent years, have a good number of citations and are published in high-quality journals, which are new works worth reading.
In addition, for more cutting-edge directions, the best starting point is to read authoritative review papers.
Each discipline basically has its own most authoritative review journal, which is small but precise, and it is bound to be found by turning over the relevant papers in recent years.
Choose the core papers you need to focus on. When you don’t know what to focus on, it is recommended to start with review papers. Look at the review papers written by academic experts and see their bird’s eye view of a whole discipline (rather than a single experiment).
However, it is easy to get lost in the sea of review papers.
Indeed, a good review often summarizes hundreds of papers, especially in small areas of research, and you need to review all the papers mentioned in the article.
But beginners don’t have to focus on details, and you can’t grasp so many details.
It should be said that intensive reading of a few core papers and fast reading of other general papers can help us understand the development of the field. How
exactly to do it?
Let’s read this hypothetical passage first: a few lines of text directly contain dozens of papers. Should we download all the papers of the above authors and read them intensively?
Don’t worry! Most of the above papers can be turned over (the so-called speed, grasp the general, remember the discovery), and you need to focus on reading through the following: (1) The first paper in a small field: Author1’s first paper determines the tone of the field, and sometimes determines the inherent problems in the field.
Take a real example: Mishkin & & Ungerleider, 1982, this paper basically established our understanding of the neural mechanism of the visual pathway, but it roughly divided the two visual pathways, and now researchers have found that the two pathways are not directly distinct through a large number of experiments.
(2) Papers published in good journals: for example, the Annual Review papers of Author22. Of course, you also need to look for papers from other top journals in the references after the paper. Papers published in top journals must be multidisciplinary and have a very high vision.
For example, the papers of AnnualReview must be created by Daniel when he is invited, and the experimental papers of other top journals should be subversive across disciplines and theories.
To a certain extent, we can even say that most of the papers are guided by the top journal papers to improve the theoretical framework.
Let’s take a look at some other details: (3) Author12 tried to overcome the public opinion, summed up different views and put forward different assumptions.
Being able to oppose other people’s views and discover new theories and methods must be unique, not to mention that it has been supported by many scholars.
This kind of paper needs to be focused on and read intensively.
(4) The research of Author23 is introduced separately. The core is to integrate different theories, which must be read carefully. In particular, we should pay attention to how the C theory integrates the differences between the former two and how to analyze the data that may be contradictory.
If the field mentioned in the article is your research direction, each of the above papers from Author1 to Author23 should be read intensively.
But the priorities between them (the order and the priority of thinking) are certainly different, and the most important thing is to summarize the important articles.
Therefore, for the first time in a field, it is advisable to read a few reviews to get a general idea, starting with reviews rather than experimental papers; secondly, take excellent review papers as clues, and select key experimental papers that have been repeatedly mentioned for intensive reading; Finally, according to the above method of repeated screening, targeted, first understand the general, in mind to build an academic framework, and then to find those repeatedly mentioned good papers to supplement the framework, knowledge is thus established.
Do a good job of intensive reading. We have just mentioned the concept of “intensive reading” repeatedly, so how can it be regarded as intensive reading?
After all, papers can not be remembered by turning them over, but need to use a set of reading methods to extract core knowledge from paper to your mind.
In my opinion, the focus of intensive reading of papers is not only reading this one, but also how to internalize knowledge points into the knowledge system in mind.
So first lay out the theoretical framework (how to build the framework can be read in the above part), and then put these papers into the framework as the cornerstone.
Therefore, the conclusion of each paper should not appear independently in your knowledge system, but should become a link in your knowledge network.
So at the beginning, don’t waste time on ordinary papers, because they are not good materials for building knowledge systems.
The collation of the literature should focus on the key points rather than paying too much attention to the details, and then turn back to the details when doing the experiment. When
I was just studying for a PhD, the first thing my boss emphasized was to summarize the literature of an experimental paper every week, in order to urge me to read in depth, and at the same time to feedback my reading depth and understanding accuracy according to my notes.
You can also practice at the beginning, just a few hundred words, and the summary of each experiment only needs to cover three aspects: why they did the research (previous theoretical development and current shortcomings), how they did the research (experimental methods, experimental hypotheses, especially the differences from the predecessors), and what they found (summary results, and what they found). If these three questions can be summed up accurately, then you really understand this paper.
And this note is also the material for future writing.
At the beginning of writing the paper, everyone has a poor memory of the literature, and it is difficult to grasp the importance, so when you take out this summary when writing the paper, you can quickly recall the content, and you can directly apply it to your paper.
In order to establish a complete subject system knowledge, choose their own field, and understand the academic trends, it is necessary to read more papers, but reading is only a prelude to scientific research, scientific research time is not unlimited, no one can spend all the time reading literature, otherwise how can we have time to experiment and write?
In my opinion, a master’s degree student who has just set foot in scientific research needs to accumulate about 100 papers before he can have a clear grasp of each small research direction in the field, so that he can judge the hot spots of research and complete certain scientific research tasks independently; while a doctor’s degree student who wants to graduate needs to increase the number to 500.
After all, the average doctoral thesis needs to cite the results of about 200 papers (estimated by the amount of a literature review plus three experimental papers), but reading only 200 papers is certainly not enough, after all, it can not be said that eating three steamed buns is full, thanks to the third steamed bun. In
order to read such a large amount of literature in the limited time of scientific research, there are several abilities that must be improved: (1) Professional foreign language reading ability The reading ability of foreign literature is not equal to simple English ability.
English ability is the basis of scientific research reading, but its requirements are more specific. Students who have
passed CET-6 have mastered basic English grammar and certain scientific research vocabulary, and the length and thoroughness of scientific research papers exceed the “reading comprehension” on the examination papers, but they will no longer deduct such details as voice, cloze test and preposition collocation.
Therefore, professional foreign literature reading requires students to master: long sentence reading ability, scientific research vocabulary, the latter is the basis of the former, and long sentence reading especially requires the ability to extract the core of the sentence (such as subject-predicate object).
Encounter new words, look up the dictionary is necessary, but must look up the English dictionary, accurately understand the meaning and usage of words, understand the vocabulary of science and technology need to read other papers, cross-mastery.
But don’t use translation software to translate into paragraphs, which will not only fail to give you accurate meaning (especially when it comes to scientific vocabulary), but also destroy your ability to read long sentences, thus affecting your overall understanding of the context of the whole paper.
I have seen quite a few graduate students who use translation software in their first year of graduate school, and they have to use software in their third year of graduate school. They read papers all day but can’t understand them.
To put it bluntly, if you use too much assistance, you will lose the ability to read a paper completely after a long time.
(2) The ability of deep reading and thinking If your reading level of literature has allowed you to read through the full text, what you need to do is to improve the efficiency of knowledge absorption, which requires our ability of deep reading and thinking. There are two levels to
improve the efficiency of knowledge absorption: the first level is to read faster, which means reading more papers per unit time, that is, “learning to read intensively.” The second level is to read better papers, that is, to spend the limited time on more critical papers, that is, “looking for core papers.” We will discuss these two points in detail later.
See here, you will also find that paper reading is not so unattainable.
It is indeed not a simple job, but a collection of necessary skills for scientific research.
As long as you can master every knowledge point on the way to reading papers, reading papers will change from “challenge” to “source of knowledge”.
. Focus on not getting lost ~
