This is a series of three articles per year featuring top researchers with an IE score of over 7, showing how to conduct research and write papers

Thread Content

Repost: I. Four essential books for postgraduate students. As the saying goes, a good memory is not as reliable as a written record; therefore, it’s important to develop the habit of taking notes from the start! As a graduate student, the following notebooks are essential: 1. Experiment record book (including the trial preparation book); this is of course the most important and necessary one, so I won’t go into further detail ；   2. Use an idea notebook: each time you read literature, note down anything that seems useful to you. Don’t overlook the ideas that arise from this; they are essential for research. A written record is better than a poor memory, as it will help generate more ideas when you review it later ；   3. A notebook for recording professional concepts and theoretical advancements – it’s impossible for anyone to know everything about the concepts in their field, especially beginners; at such times, a small notebook can be very useful ；   4. Lecture notes – this notebook may contain various miscellaneous entries; it should be used to record what was heard, as well as sudden insights and points that are not understood. These should not be overlooked!    These four books are essential for you, but for us postgraduate students who are not majoring in English, there is another notebook that we should have: a notebook for recording good English sentences. II. Key points for writing a paper 1. Choose a narrow topic and explore it in depth ； Avoid topics that are large in scope but lack substance. 2. Before writing, one should read good books, consult a large amount of materials, and focus on building academic knowledge; in this process, it is also important to make use of the Internet, especially various professional databases ；   3. The three principles of \"new topics, new methods, and new data\" (taught by the boss) ；    4. “Tackling new questions in new ways” and “making a big deal out of small matters”; in short, these are just some casual thoughts put into writing. III. How to Write Experimental Research Papers (Tang Changshu) – The Awareness of Publishing Papers: Ways to Present Basic Research Results ； Is there a rush to publish (the relationship between innovation and rigor)? ； Differences between published papers and theses (reflecting scientific facts rather than the author’s skill level) Paper formats: original articles, letters, brief reports, abstracts. Different from textbooks and lecture notes, and even more different from work summaries. Preparatory work before writing: Review and gather relevant literature; re-examine the objectives of the experiment (academic idea); ensure that all experimental data are complete and undergo another review. 1. Introduction: Raising of the issue ； Current status and background of the research ； Previous work experience ； The purpose of this work ； Approach (hypotheses can be proposed) ； object ； Methods ； Results. On the … model, observe the … indicators to explore … (purpose).      2. M & M  ⑴ The way materials are written and their meaning; **.  ⑵ Programs and indicators. Operation procedure: Sequential feasibility, operability ； Method: Ranking using a multi-criteria approach ； Introduction to reference; brief description of the improvements ； Which should be detailed or brief? ⑶ Statistical analysis 3. Results (1) Categorical description of indicators; avoid a mere list of facts. No analysis or explanation needed, but the reasoning process should be shown ；    ⑵ Text, figures, and tables should be relatively independent, but repetition should be avoided ；    ⑶ Avoid statistical errors: control, balance, randomness, repetition. Conversions between measurement-counting, absolute value-relative value, and specific indicators-comprehensive indicators. Blind and unblinded. Direct probability method and chi-square test for technical data ； Multiple sets of data and two sets of data ； Rank correlation and linear correlation ； Multifactorial and unifactorial analysis ； Paired data and independent sample data ； Non-normal distribution data ； Incorrect number of examples ； Parallel tubes, mixed samples ； Highlight differences (absolute value, Δ value, % change) ； Retention of significant figures for combined ×, ÷ ratios, subgrouping, etc. Statistical conclusions and professional conclusions.    4. Discussion ⑴ Background materials: Presentation of the issue ； Some basic information about this study (without straying too far from the topic) ；    ⑵ Analysis of the experimental results: the meaning of each indicator (compared with values from literature), and what the results indicate ；    ⑶ Further analysis of the mechanism behind the results: in combination with literature ；    ⑷ The significance of this work, conclusion or summary, and further new questions raised.    Other points to note: ① Too many cited discussion papers (unlike a dissertation), which obscure the contributions of this work ；    ② The analysis is illogical and the conclusions are inappropriate ；    ③ The discussion is too superficial, and there is a lack of familiarity with relevant literature ；    ④ Written as a work summary, it lacks academic depth ；    ⑤ Abbreviations should be used correctly, especially group abbreviations.    5. References: Why include references? ⑴ References for supporting arguments: recent and authoritative sources; preprints or abstracts should not be used ；    ⑵ Self-citation for one’s own work: Work continuity ；    ⑶ Comparison of experimental results with literature: new findings, available in brief reports, conference materials, and individual consultations ；    ⑷ Methodology: Classic literature; pay attention to the accuracy of citations, and do not quote indirectly.    6. Abstract: Introduction to the problem (Background) ； Purpose of this work ； object ； Methods (indicators, grouping) ； Main results (data, statistics) ； Conclusions and Perspectives ；    7. Reconsider the article title: off-topic, too long, or too short ；    8. Submission: Revise in accordance with the journal’s guidelines (keep a copy). Cite articles from this journal; avoid submitting the same manuscript to multiple journals. 9. Fatal flaws: unclear objectives; repetitive work without innovation; methodological issues that render the results unreliable. In clinical studies: proper selection of cases and controls; thorough analysis of the clinical relationships. IV. How to write a good discussion section: The discussion part of a scientific paper needs to be structured. It is recommended to use the following structure for the discussion section: present the main findings, the strengths and weaknesses of this study, as well as its strengths and weaknesses when compared to other studies ； In particular, the differences in the results, the significance of the study, the unanswered questions, and future research directions should be discussed. At the beginning of the discussion, it is ideal to restate the main findings in one sentence. Next, the strengths and weaknesses of this study will be thoroughly explained; neither can be ignored. In fact, editors and readers pay the most attention to the shortcomings of a study, which is inevitable in all medical research. Once editors and readers identify shortcomings in the research that have not been addressed by the author, their trust in the article is shaken, and they begin to wonder whether there are other weaknesses that neither they nor the author have noticed Secondly, connect this research with previous work; instead of boasting about how one’s own work is better than previous ones, compare their strengths and weaknesses. When comparing with other studies, be careful not to conceal your own shortcomings. It is important to discuss why different conclusions are reached from others; the author can freely speculate ； However, if one cannot understand why their research results differ from those of others, it is difficult to make such assumptions, nor should one assert that their own results are correct while those of others are wrong. 　　Next, one should discuss what one’s research “shows,” how to interpret the findings, and what significance they hold for clinicians or decision-makers At this moment, the author is in a dangerous situation; most editors and readers can understand the author’s caution to not go beyond the bounds of evidence. It is up to the readers to judge the significance of the research: they will be able to do so. The author can even point out that the research findings prove nothing, preventing readers from drawing excessive or unfounded conclusions. Finally, it should be specified which questions remain unanswered and what work needs to be continued. Obviously, editors and readers do not like exaggeration. In fact, the author often writes this part of the paper in a messy manner. Although it is impossible to stop an author from writing a piece full of speculation, evidence must not be ruined by such speculation. 　　The discussion section may sometimes require additional subheadings, but we believe that the structure proposed here is suitable for most research papers. Although a unified structure is difficult to achieve and may even be restricted, we believe it will reduce the overall length of the text, prevent inappropriate assumptions and repetitions, minimize reporting biases, and improve the overall quality of the reports. This hypothesis is fully testable. We welcome authors and readers of the BMJ to share their views; if appropriate, we will use structured discussions.   V. Tips on Writing English Articles My boss requires that each graduating doctor have more than 6 SCI articles; he never forces students to publish articles, but only teaches them how to conduct research. He only told me the answer to the following question once; afterward, if my actions didn’t match it, I would get scolded. Now I’ve grown used to thinking and acting this way, and it really does yield twice the result with half the effort. 1. Before you started your research, did you think about whether the results could be published? Where to send it? 2. Experts at writing articles first outline the structure of the article, leaving space for data; once the experiments are completed and the gaps filled in, the article can be published. It is precisely because there are rifts in the heart. 3. If you’re not sure what to write, where to post it, or what makes your work stand out compared to that of your peers, don’t start working on it yet. Go look at the literature, think about it. If you can’t figure it out, just do it; it’s better than not doing anything at all. To do this, one has to look at the literature first, right? You need to know how to structure an article, understand how others discuss the topic, and determine whether your own data shows something different from what others have found or something that hasn’t been explored by others. This process involves reading literature and thinking through things; once these are clarified, writing becomes simpler. If you do the work first and then find that someone else has already done it, or that it cannot be explained by theory, wouldn’t that be a waste of time? VI. Skills in Writing Papers: Elements of an Excellent Paper: 1. Choosing the right topic ； 2. Appropriate entry point ； 3. Concise and clear ； 4. Explain your own contributions ； 5. Reliable/reproducible results ； 6. Reusable process ； 7. A good article structure and logical flow ； 8. Selected references. Misconceptions about excellent papers: 1. The more ideas, the better ； 2、Unrelenting pursuit of revolutionary and breakthrough results ； 3. The more complex the mathematics, theories, and formulas, the better—it shows one’s intelligence ； 4. Strive for the best, unprecedented ； 5. Demonstrate authority by citing one’s own papers extensively in the references. Conditions for writing articles: 1. It must be related to research work, and there should truly be good ideas – it shouldn’t be written just for the sake of writing ； 2. Valuable results were achieved, making contributions to the academic community ； 3. The experiment is mature and can stand scrutiny ； 4. It needs to be recorded and shared with others. Key points for writing a paper: 1. Draft an outline with 3–4 levels and revise it multiple times. 2. Start with the Introduction and review previous work. 3. State the structure of the article; do not go straight into details. 4. State the motivation and basic principles behind the work, identify potential problems, and answer them yourself. 5. Explain how your work differs from that of predecessors, and outline your contributions as well as the prospects for their practical application. 6. Finally, write the Summary and Abstract, and determine the title after careful consideration. Reviewer Check List: 1. Does the paper raise a new problem or offer a new solution to an existing one? 2. What are the main results of the paper? 3. Are the experimental results sufficient? 4. What is the technical level of the paper? 5. Does the paper evaluate the effectiveness and limitations of the proposed technology/results? 6. Is the paper written clearly enough to be readable by most researchers in this field? 7. Does the paper appropriately cite and introduce relevant historical literature? 8. Should the paper be awarded a prize? IEEE Transactions on CSVT Review form: 1. To what extent does it meet the interests of the journal’s readers? 2. Evaluation of the methods used in the paper? 3. Is the result novel? 4. Are the main results correct? 5. Is the discussion clear? 6. Is there consistency (before/after, discussion/results)? 7. Is the citation sufficient? 8. Reviewer’s opinion: (Accept / Accept after a minor revision / Reject / Reject but resubmit after a major revision / Submit to another journal). VII. Essay Writing Skills: 1. Promote yourself—explain the importance of the essay. Process: a) Issue X is important ； b) Previous works A and B have studied this issue ； c) A and B have some defects ； d) We proposed Method D ； e) Conduct experiments on D and compare it with A and B ； f) Experiments have shown that D is superior to A and B ； g) Explain why D is better, while other approaches (such as E) do not work ； h) Explain the effectiveness and limitations of D ； i) Discussion on the further development of D. Key point: j) Simplicity is of the utmost importance ； k) Avoid careless mistakes, carefully verify the results, and choose words appropriately. 2. Revise carefully. Steps: a) Spend 30% of the time thinking carefully, and 70% of the time writing a draft diligently ； b) Leave the written paper alone for a while ； c) Read the paper word by word and sentence by sentence ； d) Ask others to help with reading and editing ； e) When making revisions, examine the paper from others’ perspectives (reviewers/bosses/colleagues/proofreaders) ； f) Number of careful revisions > 3 ； Total number of modifications > 5. Key point: g) Reading one’s own paper is boring, and it’s not easy to spot errors ； h) A great deal of effort is required to achieve a slight improvement in the quality of the paper. 3. Optimize English. Steps: a) Organize the paper top-down (outline/logic/process) ； b) Use other excellent papers (especially those published in the same journal/series, as well as outstanding books) as examples ； c) Ask someone to help with reading, as well as correcting grammar and word usage ； d) Keep a record of one’s own mistakes in vocabulary and grammar, and accumulate them. Key point: e) While vocabulary and grammar are important, structure and logic are even more crucial. VIII. Example of a well-structured paper: 1. Abstract — A summary of one’s work and its contributions: a) Description of the problem ； b) Explain your solution and results. 2. Introduction —— Background and outline of the article: a) Topic X is important ； b) Previous works A and B have studied this issue ； c) A and B have some defects ； d) We proposed Method D ； e) The basic characteristics of D, compared with A and B ； f) Experiments have shown that D is superior to A and B ； g) The basic structure of the article, outline. 3. Previous Work —— Explain how you differ from previous researchers: a) Categorize the work of past researchers ； b) Provide a brief review of each significant historical achievement (one to a few sentences), making sure the review is accurate, captures the key points, and avoids ambiguity ； c) Compare it with the work one has proposed oneself ； d) Do not ignore the important work of predecessors; evaluate their work fairly and avoid being too harsh ； e) Highlight the differences between your work and that of previous researchers, and it is best to provide examples of when each applies. 4. Our Work — Describe your work, which can be divided into multiple sections: a) Explain the definitions and notations from the reader’s perspective ； b) Provide pseudocode, diagrams, and corresponding explanations of the algorithm ； c) Answer potential questions that readers might have in the form of rhetorical questions ； d) Complex and lengthy proofs and details can be placed in the appendix; what’s important here is to clearly state the problem ； e) Special cases and exceptions should be explained in footnotes. 5. Experiments —— Verifying the proposed methods and approaches: a) Reasonably designing experiments (concise experiments and detailed experimental procedures) ； b) Necessary comparisons to emphasize scientific rigor ； c) Discuss and explain the significance of the results ； d) Draw a conclusion. 6. Conclusion — Summary, prospects, and closing remarks: a) A brief and quick summary ； b) Outlook for future work ； c) End the entire text. 7. References —— Comprehensive citation of relevant and important background literature: a) Selection of citations (well-known conclusions need not be cited, while the work of others should be cited) ； b) Consistent with the preceding text. 8. Others — Acknowledgments, Appendices, Footnotes. Handling rejection: 1. Understand that it is normal for manuscripts to be rejected by international peer-reviewed journals (over 70% are rejected); maintain a positive attitude. 2. Thank the editor and reviewers for their comments and efforts. 3. Ask the deputy editor how it would be best to handle this paper (resubmit/revise and resubmit/withdraw). 4. Proceed with new research or submit it to another journal after making supplementary revisions. Frequently asked questions: 1. Is it possible to submit the same manuscript to multiple journals? Never! However, a conference paper can be resubmitted to a journal after being revised. 2. Can you suggest to the deputy editor how to handle their own paper? No, but it is possible to suggest that he not have anyone review it. 3. What should be done if no reply is received from the deputy editor by April? Write a friendly inquiry letter; don’t press too hard, and don’t approach the editor-in-chief. 4. What to do if there is a serious disagreement with the deputy editor-in-chief? You can approach the editor-in-chief, but don’t do it too often. 5. What if I am in serious disagreement with the editor-in-chief’s final decision? There’s nothing to do.