After the completion of the first paper.

This is an automatically translated post by LLM. The original post is in Chinese. If you find any translation errors, please leave a comment to help me improve the translation. Thanks!

Recently, I completed my first research paper, going through the entire process from topic selection to experiments, result analysis, and paper writing. Although I took some detours along the way, I gained new insights into the methodology of scientific research and writing. This article will outline the general process of scientific research and discuss some of the details, providing reference for future research and for those who want to publish articles.

Acknowledgments: Thanks to the teachers who guided me during the experiments and paper writing, as well as friends who discussed with me.

Why Publish a Paper?

In my opinion, the purpose of doing something directly determines whether it can be done well and sustained. Therefore, before engaging in scientific research, it is important to clarify your purpose in order to have better guidance for future work.

Firstly, publishing a paper is not the ultimate goal of research. If publishing a paper is seen as the focus of scientific research, it will not be sustainable. In my view, the core purposes of writing and publishing a paper are:

1. Summarizing and consolidating research achievements at a certain stage.
2. Sharing one's work with other researchers.

Of course, besides these two core purposes, publishing a paper inevitably comes with practical benefits (academic influence, industry salary, academic assessment, etc.). Therefore, in the process of scientific research, one needs to consider both the academic attributes and the practical benefits. Striking a balance between the two may be a satisfactory solution in today's academic environment.

However, scientific research papers are primarily academic in nature and should adhere to academic norms and ethics. The core attributes should not be abandoned for the sake of practical benefits.

Paper Writing Process

A research cycle can be roughly divided into the following parts:

1. Topic selection: Choose a problem that you are interested in and that has not been well solved.
2. Method proposal: Propose a method that you believe can better solve the problem.
3. Experimentation and recording: Conduct experiments based on your method and record the results.
4. Analysis and organization of results: Analyze the experimental results. If there are shortcomings, go back to step 2 and improve the method. Otherwise, proceed to step 5.
5. Paper writing: Complete the research paper and share your method and results.

The general direction of topic selection is usually fixed, but it is important to determine the specific research questions. This step requires extensive reading of relevant literature. There are two purposes: ① Understand what others have done and have not done, and have a general grasp of the current research progress. ② Reflect and select the problems you want to solve while reading the literature. In a nutshell, "Learning without thinking is labor lost; thinking without learning is perilous." Only reading literature without thinking will not yield valuable questions, while only thinking without reading literature may lead to researching a problem that has already been solved. When I completed my first work, I did not read enough relevant literature, which led to some minor issues in the paper writing process.

When proposing a method, it is important to consider both innovation and practical significance. If the extension from "stir-fried tomatoes and eggs" to "stir-fried cherry tomatoes and quail eggs" only focuses on the taste and nutritional value, it may not be considered innovative, and it would be meaningless. However, if the research explores the scalability of cooking methods in different forms and provides insights for future cooking, it would be highly valuable. In addition, the proposal of a good method usually requires an iterative process of "proposal -> experiment -> problems -> re-proposal" to continuously refine and improve the method. Therefore, one should not be discouraged by the initial poor performance of the method, but rather repeat the above iteration process quickly to enhance the effectiveness of the method.

Experimentation and recording usually take up the largest portion of time, and this part of the process has a lower limit due to objective constraints and no upper limit influenced by subjective initiative. The experimental part does not need to be elaborated on, as it should be the basic skill of every researcher. Here, I would like to emphasize the management of experiments and the recording of results. In my opinion, good experiment management and result recording should have the following characteristics:

• Easy traceability: When you want to retrieve and view specific experimental results, you can quickly and accurately find them.
• Reproducibility: This refers not only to reproducibility in the sense of academic ethics, but also to recording the experimental configuration (including code versions, parameters, etc.) at that time, ensuring that the experiment can be reproduced quickly by retracing the experimental state at that time.
• Multiple backups: Experimental data is hard-earned, and if data is lost due to unforeseen circumstances, it can be devastating. Therefore, it is necessary to make multiple backups to ensure data security.

Result analysis is crucial and should be given due attention. Experimental data only represents certain facts, and it is through analysis and interpretation of these facts that meaningful patterns can be discovered. Various visualization methods can be used to make the hidden patterns in the data more intuitive. In addition, I personally like to pay attention to outliers generated during the experimental process and analyze the reasons behind them. Behind these anomalies, there are usually things we have not noticed before (experimental issues, unfamiliar theories, or even new discoveries...).

Finally, there is the paper writing stage. To be honest, I don't have much experience, and it is already a challenge for me to complete each section of the content. If I become more proficient in the future, I can discuss this part further.