Please enable JavaScript.

Coggle requires JavaScript to display documents.

Descriptive statistics - Coggle Diagram

- - - - Bar chart
      - Pie chart
    - - Histogram
      - Frequency polygon and relative frequency polygon
        
        在histogram的基础上找中间点连线
      - Cumulative relative frequency distribution / Ogives
        
        累计数字连线
    - - Time series chart
    - - Nominal data
        
        Cross-classification table
        
        Bar chart
      - Numerical data
        
        Scatter plot
  - - - Mean
        
        Population
        
        Sample
        
        It's sensitive to extreme values
      - Median
        
        It's unaffected by extreme values, hence it's a robust statistic
      - Mode
        
        Make no sense in continuous data
        
        It's suitable for discrete and qualitative data（离散和定性数据）
        
        Classification
        
        Unimodal
        
        bimodal
        
        Multimodal
    - - Percentiles 100
      - Deciles 10
      - Quintiles 5
      - Quartiles 4
        
        location or position
        
        even number
    - - Range
        
        Maximum - Minimum
      - Interquartile range (IQR) 四分位距
        
        Q3-Q1
        
        A robust measurement of range that is not affected by outliers
        离群值
      - Variance and standard deviation 方差，标准差
        
        Variance （e.g. 5,6,10）
        
        population
        
        Mean=(5+6+10)/3
        
        [(5-7)^2+(6-7)^2+(10-7)^2] / 3
        
        sample
        
        Mean=(5+6+10)/3
        
        [(5-7)^2+(6-7)^2+(10-7)^2] / (3-1)
        
        alternative formulate
        
        Standard deviation
        
        方差开根号
      - Coefficient of variation （CV）变异系数
      - Standardization
        
        [observation(sample里的其中一个数)-mean] / standard divation
    - - 也叫68-95-99.7法则，即以μ为中心，落在μ±σ的概率为68%，落在μ±2σ的概率为95%，落在μ±3σ的概率为99.7%
    - - covariance协方差
        
        Population
        
        sample
        
        alternative fomulate
        
        x,y same direction: positive covariance; x,y opposite direction: negative covariance
      - correlation相关系数
        
        population
        
        sample
        
        [-1, 1]
        
        r → +1 stronger positive correlation
        
        r → - 1 stronger negative correlation
        
        r → 0 weaker correlation
        
        协方差 / (x standard divation*y standard divation)
- - - - Gender, Nationality, Color
        
        no meaning, without ranking
    - - Students' grade
        
        With ranking
    - - Temperature, Length, Weight, Age
        
        Real data
  - - - At the same point of time
        
        the 2016 Australian Census
    - - In time order
        
        the ASX 200 stock market index over time
- - - - All subjects that are interested
    - - E.G. The height of QBUS5002 students
    - - A subset of subjects that are taken from a sampled population
        
        E.G. Pick 10 of QBUS5002 students randomly