Summary: Eec Lecture Slides

Read the summary and the most important questions on EEC Lecture slides

• 1 Week 1: Introduction, potential outcomes and causality

• 1.1.1.1 Correlation

  This is a preview. There are 2 more flashcards available for chapter 1.1.1.1
  Show more cards here

• Provide the 3 scenarios that would explain correlation between two variable D and Y

  1. Causal effect of D on Y
  2. Causal effect of Y on D
  3. Omitted variables: Z affects both D and Y

• 1.1.2 Potential outcome model

  This is a preview. There are 4 more flashcards available for chapter 1.1.2
  Show more cards here

• What are the 3 scenarios where ATET = ATE, and why?

  1) All students participate in treatment.
  ATE concerns the entire population, whereas ATET only the treated.
  If all students participate, then "the treated" = ''the population"
  
  2) All participants benefitted the same from treatment.
  I cannot see yet why. 
  
  3) Assignment to treatment is randomized
  I think the why here is the same with number 1; your entire population is now considered "the treated"

• 1.2.2.1 Example linear regression model - adding control variables

  This is a preview. There are 2 more flashcards available for chapter 1.2.2.1
  Show more cards here

• What do we need for OLS to provide unbiased and consistent estimators for our (control) variables

  We need: zero conditional mean assumption to be true:
  (E[Ui |Di , Xi ] = 0)

• If your selection (of participants) is based on certain characteristic Xi, do you need to include Xi as a control variable? And if so, why?

  Yes, to ensure consistency and unbiasedness of the estimator in question: δ

• 1.2.2.2 Intermediate outcomes: Example

• Treatment: banning the sale of soda and sweets in school canteens. Interested in the effect of the treatment on pupil’s health (measured as BMI 6 months later). Sample of 200 high schools in NL: 50% randomly treatedFor each of the following, indicate whether it is an intermediate outcome or not: A) GenderB) BMI three months after start of the treatment.  C) BMI before the start of the treatment.D) Euros spent on food items (after treatment)

  I would say B and C, double check with teacher.

• 1.2.2.3 Heterogeneous treatment effects

  This is a preview. There are 2 more flashcards available for chapter 1.2.2.3
  Show more cards here

• Give a short definition of Heterogeneous treatment effects

  Different individuals responding differently to treatment, due to different characteristics

• What is one recommend course of action when you are dealing with heterogeneous treatment effects?

  1. Group* the individuals with particular characteristics, denoted with X, together
  2. Apply the differences in means estimator
  
  * they call it stratify

• When can the recommended course of action (asked in a previous question) be problematic?

  1. When the characteristic variable X is continuous
  2. When the stratified samples (aka groups by characteristics) become too small.

• 3 Week 3

  This is a preview. There are 19 more flashcards available for chapter 3
  Show more cards here

• Name the 4 types of experiments from the categories Field, and, Experiment

  1. Laboratory experiments
  2. Artefactual field experiment
  3. Framed field experiment
  4. Natural field experiment

• Provide the characteristics for: Laboratory Experiments

  Stylized environment
  stylized context
  usual participants