Chapter 11 Flashcards
correlation
When x is higher y is higher or lower
causation
X causes u
If we can somehow increase x, y will increase or decrease
reverse causation
Y→ X
spurious correlation
Z → X Z→ Y
Relationship of z cancels out relationship between x and y
confounding
Z - When removes relationship to not be casual
Example 1: as ice cream sales increase in a city, the cities murder rate increases as well
X → ice cream Y → homicide rates what could Z be? - Heat, example of spurious relationship
four hurdles of establishing causality
Is there a credible causal mechanism that connects X to Y?
Can we rule out the possibility that Y could cause X?
Is there covariation between X and Y?
Have we controlled for all possible confounding variables?
antecedent
Causally comes before both independent and dependent variables
simultaneity bias
Stems from causal relationships that simultaneously go in both directions
causal mechanisms
Some notion of how independent and dependent variables are related
indirect effect
Sense that independent variable influences dependent only through causal effect on onore or more intervening variables
direct effect
When intervening causal process or mechanism is nearly immediate, fairly mechanical, or so obvious that it does not need spelling out
self-selection
Influence of choices
Individuals studied choose the category of level of independent variable of interest
endogeneity
Independent variables value caused by variables or processes that also effect dependent variable
Often due to self-selection
intervention / treatment
Focused attempt to manipulate specific features of environment or context in a way that is deliberately independent of and apart from the choices, preferences or other characteristics of subjects in experiment
experimental control
Holding constant
