Week 2 Lecture Flashcards
List the 5 types of explanation
1: Deductive nomological/covering laws
2: Probabilistic-nomological explanation
3: Functional explanations
4: Intentional explanations
5: Mechanistic explanations
Define deductive-nomological/covering laws
laws that posit strong, deterministic relationships; rare in social sciences
example: Rich people vote for the Democrats –> John is rich –> therefore John votes/voted/will vote for the Democrats
Define probabilistic-nomological explanation
similar to covering laws, but allows for probabilistic relationships
example: rich people tend to vote for the Democrats –> John is rich –> therefore John is probably going to vote Democrat
Define functional explanations
explain the causes of behavior by its consequences (teleologically); look at the system level and try to determine why certain things prevail and the effects that characteristics have on the function of that system
example: John believes its in his best interest to vote Democrat
Define intentional explanations
explain behavior by beliefs and intentions, such as preferences or reasons
Define mechanistic explanations
explain by identifying the causal paths linking causes and effects; trace the links of a causal chain or the interactions of a mechanistic model; identify a sequence of relevant events
Define causality
what would happen (or would have happened) if X changes but everything else stays the same –> counterfactual definition
Define cause
a cause is something that is a difference maker for an outcome
What does it mean that the definition of causality is “counterfactual”?
causality is unobservable; we only experience one version of the world and we cannot rewind and try what would happen if we changed things
What is the difference between correlation and association
Correlation: one specific measure of associations
Association: whether or not two things go together
Name and describe the 5 different reasons for two variables to be associated
1: Chance –> randomness
2: Causation –> X causes Y
3: Reverse causation –> Y causes X
4: Confounder bias/omitted variable –> Z causes X and Y
5: Collider bias –> conditioned on a shared effect of X and Y
What are the four types of variables?
1: Confounder (X <– Z –> Y)
2: Mediator (X –> M –> Y)
3: Collider (X –> Y <– U)
4: Instrument (I –> X, but not X –> I)
What are the five essential features of the logic of experiments
1: random selection
2: pre-test
3: treatment and control groups
4: random assignment
5: post-test
