BlackCat

The Effect

Book Reading 2 min read

the effect — GPT-generated monochrome cover

URL：https://theeffectbook.net

Designing Reasearch

1.1

Research Question

Research Question : try to answer/plan to answer by research

good research Question:

  • well-defined 清晰定义 (clear)
  • answerable 有方法能够回答
  • understandable 对他人容易理解的 (知道你在问什么)
flowchart TD
    A["Research Question : try to answer/plan to answer by research"]
    A --> B["well-defined 清晰定义 (clear)"]
    A --> C["answerable 有方法能够回答"]
    A --> D["understandable 对他人容易理解的 (知道你在问什么)"]

1.2 Empirical Research

书籍主要关于 quantitative empirical research

  • (由数据驱动的实验研究)

Empirical research

Empirical research: 使用结构化观察（来自真实世界）去回答研究问题

Emprical(based on observations from the real world)

举例： 交通问题

不需要对凭空推理司机会做什么

观察司机会做什么 （例如问卷）

flowchart LR
    A["不需要对凭空推理司机会做什么"]
    B["观察司机会做什么 （例如问卷）"]

Quantitative empirical research

  • Empirical research using quantitative measurement
  • 更多的data sets
  • 更少的interviews

quantitative empirical research 很难

  • 数据

    • hard to take precisely (量度)
    • hard to interpret accurately(对数据的理解)

  • 统计学

    • 很难(内容 知识)

“One particularly sticky problem with quantitative empirical research is that the numbers that we observe often don’t tell us exactly what we want to know.”

data 是 raw 的，可以回答目前的状态， 无法自动回答我们的研究问题

flowchart TD
    A["quantitative empirical research  很难"]
    A --> B["数据"]
    B --> C["hard to take precisely (量度)"]
    B --> D["hard to interpret accurately"]
    A --> E["统计学"]
    E --> F["很难(内容 知识)"]

1.3 Why Research Needs a design

"looking at what happens when it’s not."

Bad Research Design

flowchart LR
    A["Bad Research Design"] --> B["Can not answer the original questions"]

Bad Research Design -> Can not answer the original questions

透过良好设计的程序 尽量杜绝错误的发生 例如做题时候的data verfication 大幅降低careless的错误率

Keep reading

View all