Business Statistics — Lecture 2 (商业统计 第2讲)
1. Data, Elements, Variables, Observations (数据、要素、变量、观测值)
Explanation / 解释
- Data are fact-based information (数字、图表、表格) for collection, analysis, interpretation. (数据是基于事实的信息，用于收集、分析和解释)
- Elements are the entities from which data are collected. (要素是收集数据的对象/实体)
- Variables are characteristics of elements. (变量是要素的特征/属性)
- Observations are sets of measurements for each element. (观测值是针对要素的测量数据集合)
Example / 例子
- Data: Company sales = $500,000 in Jan. (一月公司销售额 = 50万美元)
- Element: Each student in a class. (每个学生)
- Variable: Age, gender, GPA. (年龄、性别、绩点)
- Observation: {Age=20, Gender=F, GPA=3.5}. (观测值记录在行里)
Extension / 拓展
- Data are the foundation of decision-making. (数据是决策的基础)
- Identifying elements ensures representativeness. (识别要素保证代表性)
- Variables can be qualitative or quantitative. (变量可定性/定量)
- Observations are stored in dataset rows. (观测值存放在数据行)
2. Scales of Measurement (测量尺度)
Explanation / 解释
- Four scales: nominal, ordinal, interval, ratio. (四类尺度：名义、顺序、区间、比率)
Example / 例子
- Nominal: Gender (Male/Female). (名义：性别 男/女)
- Ordinal: Rating (Very good=5, Very bad=1). (顺序：评分 非常好=5, 非常差=1)
- Interval: Temperature (°C). (区间：温度 °C)
- Ratio: Distance, probability. (比率：距离、概率)
Extension / 拓展
- Correct scale selection impacts valid analysis. (正确的尺度选择决定有效分析方法)
3. Data Types (数据类型)
Explanation / 解释
- Categorical (Nominal/Ordinal). (类别数据：名义/顺序)
- Quantitative (Interval/Ratio). (定量数据：区间/比率)
Example / 例子
- Categorical: Male=1, Female=2; Very good=5, Very bad=1.
(类别：男=1, 女=2; 非常好=5, 非常差=1) - Quantitative: Distance 10km vs. 3km; Loss chance 0.5 vs. 0.3.
(定量：距离10公里 vs. 3公里；亏损概率0.5 vs. 0.3)
Extension / 拓展
- Categorical: analyzed with counts, percentages. (类别数据用频数/百分比分析)
- Quantitative: supports mean, variance, regression. (定量数据支持均值、方差、回归分析)
4. Data Collection Methods (数据收集方式)
Explanation / 解释
- Cross-sectional: data at one point in time. (横截面数据：某一时间点收集)
- Time series: data across periods. (时间序列：多个时间段收集)
Example / 例子
- Cross-sectional: May 2025 customer survey. (横截面：2025年5月顾客调查)
- Time series: Sales from 2020–2025. (时间序列：2020–2025年销售额)
Extension / 拓展
- Cross-sectional: good for group comparison. (横截面对比群体)
- Time series: useful for trend, forecasting. (时间序列用于趋势和预测)
5. Data Sources (数据来源)
Explanation / 解释
- Existing sources (internal/external). (现有来源：内部/外部)
- Statistical studies (experimental/observational). (统计研究：实验/观察)
Example / 例子
- Internal: sales records. (内部：销售记录)
- External: industry reports. (外部：行业报告)
Extension / 拓展
- Sources differ in cost, reliability, accessibility. (来源在成本、可靠性、可获得性上不同)
6. Data Acquisition Errors (数据获取误差)
Explanation / 解释
- Errors in collection can be reduced by consistency checks/common sense.
(数据收集中的误差可用一致性检查和常识减少)
Example / 例子
- Reject outlier: Age = 200. (排除异常值：年龄=200)
Extension / 拓展
- Data cleaning is essential before analysis. (分析前必须进行数据清理)