[跟林師模教授學 Quantitative Methods 的筆記]

需再複習有關「卡方檢定」，還有「同質性檢定、適合度檢定、獨立性檢定 ....」等等的差別，所以準備重新dive in



Statistics Testing

  • 統計學上使用到的各種「檢定」方法，可以根據不同的目的和資料類型來分類。以下是一些常見的「檢定」方法的分類和例子，並附上相關的網頁連結：
    目的 資料類型 檢定方法 連結
    比較兩個或多個群組之間的平均數差異 連續變數 t檢定、ANOVA、MANOVA等 1
    比較兩個或多個群組之間的比例差異 二元變數 z檢定、卡方檢定等 2
    比較兩個或多個群組之間的中位數差異 順序變數 曼-惠特尼U檢定、克魯斯卡爾-華利斯檢定等 3
    測試一個或多個變數之間的相關性 連續變數或順序變數 皮爾森相關係數、斯皮爾曼相關係數等 4
    測試一個或多個變數之間的因果關係 連續變數或二元變數 迴歸分析、邏輯斯迴歸等 [5]
    以上只是一些常見的「檢定」方法，並不是全部。在選擇「檢定」方法時，還需要考慮資料是否符合一些基本假設，例如獨立性、常態性、同質性等。如果資料不符合這些假設，則需要使用非參數檢定或轉換資料。

  • 2.查看常用25種統計方法
  • 3.找Youtube幫幫忙
  • 4.總是被這一堆術語搞得頭昏腦脹。用哪種統計分析方式好？(卡方,Anova,T-test, or regression?)
  • 5.假設檢定的步驟和範例
    虛無假設顯著性檢定(null hypothesis significance testing)或簡單稱為假設檢定(hypothesis testing)，是一種運用統計模型來檢驗研究問題的方式，也是用來測試研究人員操縱的變項，也就是(自變項)是否具有效果的同濟推動方法可以說是統計分析中最關鍵也相當繁複的一種過程。
  • 6.數據分析師必看！9大統計法整理，這篇讓你找到最適合的統計分析方式！
  • 7.統計學知識入口網站JMP什麼是 t 檢定？免費的線上統計學介紹

  • Appendix 檢定的「目的」及諸多不同檢定方法的定義....

    What is the purpose of the various "test" methods used in statistics? How many testing methods are there? Please describe the definition of each test. Please list the above information in a markdown table:

    The purpose of various statistical tests is to validate hypotheses, quantify uncertainty, and draw meaningful conclusions from data.
    These tests help determine whether observed patterns in data reflect true relationships or are due to chance.
    They provide mathematical rigor, enable reliable inferences, guide predictive model development, facilitate comparisons over time, support objective decision-making, and enhance reproducibility in research.

    There are numerous statistical testing methods, each suited for different types of data and research questions. Below is a table summarizing some common statistical tests along with their definitions:

    Statistical Test Definition.


    🍉(談到:兩個卡方分配=F分配，可參考蘇志雄介紹重要的機率分配：Z分配、卡方分配、t分配與F分配會更清楚。)
    🍉看两万一张饼讲解统计学基础知识
    😊什么是概率分布？｜Probability Distributions in Stats 知道概率分佈的道理，不一定讓你能預測但應該可以助你避免犯錯，讓你學會「未知全貌、不予置評」的道理，不會跟著小樣本的走向而起舞。
    😊什么是t分布？｜T Distribution in Stats Clearly Explained
    🍉zedstatistics testing 介紹各種testing methods

    t分布（Student's t-distribution）在都柏林Arthur Guinness吉尼斯父子釀酒廠從事實驗和數據分析工作的William Sealy Gosset，1908年在Biometrika雜誌以筆名Student發表的論文：均值的或然誤差（The Probable Error of a Mean）。文中提出了t分布的存在，打破了常態分布一統天下的局面，開創了小樣本統計推論的新紀元，在統計學史上具有劃時代的意義。

    ▼1 Bing: 比較Web1.0和Web2.0和Web3.0的差別

    比較Web1.0和Web2.0和Web3.0的差別


    ▼2 A master's Thesis

    演講:怎樣寫好碩士論文
    2023/10/25-402B
    中原大學 財金系陳若暉
    著作:     金融科技(附習題與解答)：金融與數位技術的跨界整合
    | 人工智慧：Python程式應用
    | AI人工智慧導論：理論、實務及素養

    A master's Thesis
    -Is the final project for a master's degree
    -Describes the results of original researsh
    -Contains a thesis question and an answer to it


    Think well about your topic
    -Try to choose that in intersting for you
    -Academic interest favourifle subject of study

    Narrow down the topic
    -Define the exact boundaries of your work.
    Indentify your research object
    -try to outline the period, cycle, or course that your paper's practical part will or could thouch upon
    Think about the title of your work

    1.Think about the ojectives of writing a thesis
    1.to get degree
    2.enoy the work
    3.get a job afterward
    4.to be useful

    2.Generate thesis ideas
    1.thinking about your field, find the gap, what new analyses can you offer?
    get data(50% of your work)model + methodolory + conclusion + runing your + (conclusion: contribution)
    2.think your favorite subject of study-authors,theory ,time period,
    3. consider skimming through papers see if there is any apparent. read
    future study, they provide something he can not finish.
    4. favorite professors

    3.Choose the right topic
    1.find best fit the oj=bjectiver from the first stip
    2.
    3.
    4.
    5.Counduct your research
    6.Choose your committee members. 3 professors

    to be a Master: 1 thesis
    to be a phD: 3 thesis
    to be a professor: continure learning many new models

    II selecting our text
    1).complere a literature review
    2).choose your sources
    3).manage your citations
    1-keep track of your citations
    2-use the in-text citation format appropriate
    3-
    4-manage your citations consider using a citation management software such as EndNote Mendeley, or Zotero.

    III Planning an Outline
    1.know the requirement for your field/department qualitative or quantative 2.Nail down your thesis idea

    KEY TYPES and METHODS for data collection

    ETF，英文全名為Exchanged Traded Fund，中文為指數股票型基金，其透過追蹤、模擬或複製標的指數之績效表現為投資目標，進行一籃子地股票投資。

    collect data is use a lot time, you need to make it efficient
    Steps to carry out th Delphi Method
    Web Scraping 用Python從網路就可抓到資料

    Master's Thesis outline
    know what to include: introduction body conclusion

    --Moving through the Writing Process------set up working schedule 1.Make a schedule if you know how much time you have to complete the project and break it up into manageable parts with individual due dates. 2.write a little every day. Your exact timeframe will differ, but you should try to give yourself as much time as possible to write. 找資料的最好兩個工具-圖書館- 7.[EconLit-EBSCOhost]
    3.Web of Science-[WOS SCIE / SSCI]

    1.what is the contribution?
    2.frmework method and data
    3.sources of Data
    4.describe some important and/or interesting empirical results -4.emperical study
    5.Can you think of a way to build upon this work? describe. -feature study.

    -Key parper you may pay time to read it otherwise don't spent too much time. Gingramer
    Ginger writer [Grammarly free gramar checker] [Turnitin]
    Make sure you abide by these guidelines to avoid any potential setbacks during this final stage.
    [Pro Quest] ProQuest有限責任公司（英語：ProQuest LLC）是一家位於美國密西根州安娜堡的全球性資訊內容和技術企業，由尤金·鮑爾於1938年以「大學微縮膠捲」（University Microfilms）之名建立。 該公司業務範圍包涵為圖書館提供解決方案、應用程式、及其它產品 。

    Professor Achivement with SSCI publication
    SSCI > SCI
    average accumelated
    time required for submission and publiation SSCI

    no nu.of submission period months publick mth total period months. 56.65months

    發表了97篇了 SSCI有11篇;

    Alternative access to Elsevier [BEALL's LIST] ethics
    [Elsevier Alternative Access Review 2022 - UW Libraries]
    大前研一On and OFF are divided into work and leisure


    ▼3 pfZF

    ptZF

    p-value、t-value、Z-value 和 F-value 的定義及其異同：

    定義
    • p-value：p-value（概率值）是指在假設零假设为真时，获得当前或更极端数据的概率。p-value 越小，说明数据与零假设的偏离越大，从而更有可能拒绝零假设
    The p-value (probability value) is the probability of obtaining test results at least as extreme as the observed results, assuming that the null hypothesis is true. A smaller p-value indicates stronger evidence against the null hypothesis.
    1 P-Value: What It Is, How to Calculate It, and Examples
    2 Understanding P-values | Definition and Examples

    Wikipedia 3 p-value 。

    • t-value：t-value 是指样本均值与假设均值之间的差异相对于样本标准误的倍数。它用于 t 检验中，以确定样本均值是否显著不同于假设均值
    The t-value is a ratio of the departure of the estimated value of a parameter from its hypothesized value to its standard error. It is used in t-tests to determine if there is a significant difference between the means of two groups.

    4 What are T Values and P Values in Statistics?
    Wikipedia 5 t-statistic

    6 How t-Tests Work: t-Values, t-Distributions, and Probabilities 。

    • Z-value：Z-value（或 Z-score）是指数据点与总体均值之间的差异，以标准差为单位。它用于标准正态分布中，表示数据点在分布中的位置
    The Z-value (or Z-score) represents the number of standard deviations a data point is from the mean of the data set. It is used in standard normal distribution to determine the position of a data point within the distribution.
    7 Z-score: Definition, Formula, and Uses
    8 Z-Score: Definition, Formula, Calculation & Interpretation
    Statistics How To 9 Z-Score: Definition, Formula and Calculation 。

    • F-value：F-value 是在方差分析（ANOVA）或回归分析中使用的统计量，用于比较组间方差与组内方差的比值。F-value 越大，说明组间差异越显著
    The F-value is a ratio of two variances. It is used in ANOVA (Analysis of Variance) to compare the variances between groups and within groups. A higher F-value indicates a greater degree of difference between group means.
    Statistics How To 10 F Statistic / F Value: Simple Definition and Interpretation

    Wikipedia 11 F-test
    12 How F-tests work in Analysis of Variance (ANOVA)

    IBM 13 F value
    14 F-value (ANOVA): Get the Most Out of Your Data Analysis 。


    • 異同
      指標 定義 用途 计算方法 适用场景
      p-value 在零假设为真时，获得当前或更极端数据的概率
      Probability of obtaining results at least as extreme as observed, assuming null hypothesis is true       		假设检验
      Hypothesis testing       		通过统计软件或 p-value 表计算
      Calculated using statistical software or p-value tables       		各种统计检验，如 t 检验、F 检验等
      Various statistical tests like t-tests, F-tests, etc.
      t-value 样本均值与假设均值之间的差异相对于样本标准误的倍数
      Ratio of the difference between sample mean and hypothesized mean to the standard error       		t 检验
      t-tests       		((bar{x} - μ) / (s / sqrt{n}))
      		 样本均值与假设均值的比较
      Comparing sample mean to hypothesized mean
      Z-value 数据点与总体均值之间的差异，以标准差为单位
      Number of standard deviations a data point is from the mean       		标准正态分布
      Standard normal distribution       		((x - μ) / σ)
      		 数据点在正态分布中的位置
      Position of a data point in a normal distribution
      F-value 组间方差与组内方差的比值
      Ratio of variance between groups to variance within groups       		方差分析（ANOVA）
      ANOVA (Analysis of Variance)       		(frac{text{组间方差}}{text{组内方差}})
      		 多组均值的比较
      Comparing means of multiple groups
    Key Points
    • p-value is used to determine the significance of the results.
    • t-value and Z-value are used to measure how far a sample statistic is from the population parameter.
    • F-value is used to compare variances across multiple groups.


    ▼4 aaa

    aaa


    ▼5 aaa

    aaa



    | 海闊天空 |