L3 Flashcards
causality
how change in x affect Y
if everything else is equal
casual effect of x on Y = change in Y-counterfactual change Y(would happen anyway)
correlation without casual effect because:
spurious correlation (chance) omitted variable bias (third variable affecting both) or selection bias reverse causality (opposite relation)
tree ways to identify causal effects
randomized controlled trial (RCT)
Natural experience
Instrumental variables
Explain randomized controlled trial
and the two economic games
compare group with treatment to control group
dictator game (give as you want = no reason to give) Ultimatum game (give and accept = give min amount above 0)
external validity
can result in RCT be applicable to outside world?
since people are being monitored that affects their behaviour
problem with RCT
Hawthorne effect = change behavior cause they are being studied
John Henry effect = control group try to outperform
ethical issues (no treatment for control group, waste of resource) Practical issue (expensive, spill over effects, attrition(lose people in experiment))
natural experiment
occurs naturally, not created and can be observed
environmental shock, sudden policy change, random administrative error
instrumental variables
+ 2 requirements
use an instrument variable that only affects x to see it y changes
- relevance = explain that it impacts x
- exclusion restriction = explain that it doesn’t impact Y