# Summary: Research Lab

• This + 400k other summaries
• A unique study and practice tool
• Never study anything twice again
• Get the grades you hope for
• 100% sure, 100% understanding
Remember faster, study better. Scientifically proven.
PLEASE KNOW!!! There are just 86 flashcards and notes available for this material. This summary might not be complete. Please search similar or other summaries.

## Read the summary and the most important questions on Research lab

• ### 1.1 Estimating causal relationships

• #### What is the econometric goal for estimating a causal relationship?

The selection bias (omitted variable bias) should be equal to zero

E(Y1|X=1) - E(Y0|X=1) + E(Y0|X=1) - E(Y0|X=0)
• ### 1.2 Selection bias (omitted variable bias)

• #### What solves the omitted variable bias?

Randomized control trials; it makes X independent of the observed and unobserved characteristics of the two groups

This gives us B*ATT
• ### 1.3 Regressions in a non experimental context

• #### What are the five identification strategies that use regression analysis to approximate a randomized experiment and identify a causal relationship?

1. Selection-on-observables
2. DID
3. Fixed effects
4. Instrumental variables (IV)
5. Regression discontinuity design (RDD)

• ### 2.3 OLS regressions

This is a preview. There are 2 more flashcards available for chapter 2.3

• #### What are the key assumptions for OLS?

1. Exogeneity assumption: X and the error term are uncorrelated
- Cov(x,e)=0

2. Homoskedasticity assumption: for every X the errors have the same variance

3. Independence assumption: Conditional on X, the errors are uncorrelated
E(e|x)=0
• ### 2.4 Control variables

This is a preview. There are 4 more flashcards available for chapter 2.4

• #### What are noisy estimates? (outliers)

High variance estimate

= the result of vertical outlier (error term is high or low)
• #### What is the difference between omitted variables and bad control variables?

Bad control variables are still in the regression, but they are biased.

Omitted variables make the indendent variable in the regression biased.
• ### 2.5 Predictive validity framework (Libby boxes)

• #### What are libby boxes?

Help researchers identify a causal relationship and to set up a research design
• #### What do Libby boxes exist of?

1. X conceptual: construct X
2. Y conceptual: construct y
3. X operational: variable x
4. Y operational: variable y
• ### 2.6 Irani and Oesch (2013)

This is a preview. There are 3 more flashcards available for chapter 2.6

• #### What was the research question for Irani and Oesch (2013)?

Does monitoring by analysts lead to better quality of corporate disclosure?
• #### What was the problem with the relation between analyst monitoring and corporate disclosure?

1. Simultaneous problem - X-->Y and Y--> X
2. Omitted variable bias
PLEASE KNOW!!! There are just 86 flashcards and notes available for this material. This summary might not be complete. Please search similar or other summaries.