Arthals

Arthals

bilibili
github
HomeArchivesAbout

Learning Biostatistics from a Beginner's Perspective

#Statistics#Study#R273
AI Translation
This post is translated from Chinese into English through AI.View Original
AI-generated summary
The code portion of this project is open source under the GPL3 license, while other parts are publicly available under the CC BY-NC-SA 4.0 license. The project includes contributions from the author's roommates, online sources, and a teacher from Capital University of Economics and Business. The author does not guarantee the accuracy of the content and advises users to approach it critically. The project can be found on GitHub. The author shares personal experiences and opinions about the course, expressing dissatisfaction with the teaching of coding and assignments. They highlight the lack of emphasis on practical coding skills and the disorganized and confusing content of the course. The author also mentions the challenges of managing time and keeping up with the course requirements. They share their personal learning process, including initial neglect, struggling with assignments, and eventually dedicating more effort to studying. The author provides advice, such as using R in VS Code, self-studying probability theory and mathematical statistics, and using modern software like R instead of STATA or SPSS. They also suggest taking notes during class, using Markdown and LaTeX for organizing notes and assignments, and learning regular expressions for data entry.

Copyright Statement#

All code portions of this project are open source under the GPL3 license. Other parts are publicly available under the CC BY-NC-SA 4.0 license.

This project is mainly organized by me personally. Some of the assignment code comes from my roommates, and some content comes from the internet and Professor Tong Qiang's courseware at Capital University of Economics and Business. I would like to express my gratitude to them.

I do not guarantee the correctness of the content, so please use this material with a critical mindset.

Personal Experience#

The following is an excerpt from Tree Hole #4591252. It may contain personal bias. If it is offensive, please understand:

The part of code and practical assignments in the teaching of Health Statistics is really hard to describe... For me, I don't even understand the course taught by the teacher, and most of the time, I have to rely on myself to Google/Baidu/CSDN to learn the code taught by the teaching assistant... The teaching assistant's assignments are mostly "given a piece of code, tell you that this code can produce the required results," and that's it. They didn't even teach us how to input and output properly, and then they gave us a sample assignment for us to copy (even in the case of a half-understanding of statistics), which is fine, but the most outrageous thing is that the teaching assistant will find your shortcomings in various details after you submit the assignment, and then criticize you for not having statistical literacy while deducting your assignment score. The problem is that in the first half of the semester, most students didn't have such a high level of statistical literacy, who would have thought that every sentence would require an assumption, attention to word usage, and so on, it's really outrageous.

I feel that Health Statistics is a course that does not focus on practical teaching, but still imposes some code assignments under the guise of "practice," completely ignoring the fact that everyone's coding level is not high (I consider myself to be one of the stronger ones in programming, but I still feel confused when learning code/writing assignments), and then directly teaches various functions in R language without even having a basic understanding of R language teaching. This means that you may have to face the embarrassment of "never seen before, but you are required to write it" for some commonly used syntax, and I guess some people may have to try each function prefix like q, d, p, r...

Alas, it's not really feasible to say that giving points is not enough after looking around, fortunately, it's pass/fail, so be it.

[Eve] Our class teacher said in class: This is a course on health statistics, not a programming course.

[Francis] As a person who has been through it, it is originally a course on statistical theory, not teaching you how to code.

[Grace] Programming is not important. The main problem is that this course on statistics is taught very poorly, with no organization, and the content is extremely chaotic, not suitable for beginners to learn.

The biggest problem with this course is that it is a 3-credit major course. Without pass/fail, you must complete the learning of probability theory and mathematical statistics within a limited time, and also learn some R language from scratch. In addition, you must keep up with the progress of the teacher every week, otherwise, the deduction of points in the quizzes will directly affect your GPA.

Personal Learning Process#

  1. At the beginning, I didn't pay much attention to it and was confident in my programming foundation. I thought the teacher's lectures on the normal distribution were relatively simple, and there was not much effective class time.
  2. I found that there were four hours of computer practice every week, and there were also 1-2 quizzes. If I didn't listen carefully or study on my own, it was easy to get 6/10 or 7/10.
  3. When I started doing assignments, I found that I didn't know much, so I hurriedly copied the teaching assistant's template before the deadline, resulting in a score of 80/100.
  4. Repeat the above cycle once, and feel that this course is more and more outrageous.
  5. I decided to study hard and wrote the third assignment seriously, but lost points in some details, scoring 95/100.
  6. I felt that it wasn't that difficult, so I started to slack off and couldn't keep up with the progress.
  7. During class, I heard some other students confidently answering the teacher's questions and discussing the knowledge points enthusiastically during breaks. I didn't know what they were discussing, but I felt that I could handle this course, so I didn't pay much attention to it.
  8. During class, I felt that the teacher's lectures were not systematic, and I felt the same when reading the textbook. The parts related to mathematical reasoning were often vague and skipped, and I couldn't connect the logical points of knowledge.
  9. I didn't pay attention in class, so I needed to "preview" all the knowledge points of the class before doing the assignments. Each assignment took more than 2 hours. Some assignments that required drawing charts were even more time-consuming, so I made a three-line graph HTML generator with Python because of my OCD.
  10. In the second half of the semester, I started to study this course seriously, gave up listening in class, self-studied probability theory and mathematical statistics, and then connected the knowledge points from the beginning. Although it was still vague, at least it became systematic.
  11. I reviewed overnight before the exam, spent two days summarizing notes, and reviewed past exam questions (with a high repetition rate of multiple-choice questions).
  12. During the exam, I did the multiple-choice questions carefully and tried my best to explain the definitions. Due to the pandemic, I didn't have computer-based exams, so I finished the long questions hastily. Finally, I passed as I wished.
  13. After the exam, I complained to students from other schools studying economics. He showed me their R language courseware, and I significantly felt that it was better than ours, as they were really teaching the language (although it lacked many statistical parts). I felt that I learned more from their courseware in two hours than I did in a whole semester of learning R language.

Personal Experience#

  1. I recommend configuring R language for use in modern editors like VS Code (Google keywords: R in VSCode, Radian). It provides a much better experience than R Studio, with an elegant interface and more customizable settings (such as chart output), and you can even benefit from Copilot's intelligent suggestions for free.
  2. In terms of statistical knowledge, I recommend self-studying with Professor Chen Xiru's "Probability Theory and Mathematical Statistics" as a companion. You can preview before class or read the PPT at the beginning.
  3. In terms of software, focusing on learning R is enough. You don't need to learn STATA, SPSS, etc., unless you are a GUI enthusiast and can tolerate the outdated interface design of these software.
  4. Don't take the teacher's ridicule to heart in class, and don't worry about how other students are learning.
  5. Write assignments seriously, take notes during class, and I recommend typing on the computer. OCR recognition of courseware + typesetting by yourself is not only comfortable but also truly systematic for your learning.
  6. If you feel that entering data is quite troublesome when doing assignments (usually when the teaching assistant is lazy and doesn't provide you with a CSV file), you can learn regular expressions to quickly enter data, combined with VS Code/OCR for fast data entry.
  7. If you, like me, have OCD and want to draw electronic three-line graphs, you don't have to manually create a three-line graph generator like me. You can simply use Excel/Numbers to adjust the style and then take a screenshot of a PNG to insert.
  8. When organizing your notes and writing assignments, you can also learn Markdown and LaTeX syntax. I recommend using Typora for Markdown editing, which provides real-time preview and customizable CSS styles, making writing assignments not only simple but also elegant. For LaTeX formulas, you can use Mathpix, which can recognize mathematical formulas in courseware, and students' email addresses have 100 free uses per month, which is sufficient. For detailed instructions on this part, please refer to R 语言.md.
Loading...
Ownership of this post data is guaranteed by blockchain and smart contracts to the creator alone.