Final Exam Flashcards
Participant Observant
a method of qualitative research by which the researcher ob-
serves and engages with the community or social groups from which she wants to under-
stand a social phenomenon
Belmont Report
code providing guidelines for ethical research
3 Principles of the Belmont Report
- respect (for persons) 2. Benefit 3. Justice
Natural Experiment
researcher does not randomly
assign the level of the independent variable
4 Hurdles to establishing a casual relationship
- Is there a correlation between X and y?
- Can we rule out reverse causation?
- Is there a credible causal mechanism?
- Have we controlled for all confounding variables?
Goal of sampling
to make inferences about a
population by studying a sample
representative surveys
(we make informed estimates of how close we are to the true
population value)
target population
Population to which you would like to generalize your results
sampling frame
Set of all cases from which you will select the sample.
non-response bias
a bias that occurs when a certain group of people are the ones that respond to a survey
coverage
errors
failure to cover adequately all components of the population being studied
Survey Experiments
treatment is usually textual or information, DV is something
you could ask in a survey. Often online.
Lab Experiments
takes place in a lab. Often involves behavior, interaction or
fancy measurements (e.g., skin conductance)
Field Experiments
DV is usually real-world behavior, treatments vary.
Natural experiments
treatment is not assigned by researchers
Fundamental problem of causal inference
for any individual unit, we can observe only one of Y(1) or Y(0) (controlled by limiting confounders)
Internal validity
Did we accurately identify the causal effect we care about?
Are there factors other than the intended independent vari-
able that could be responsible for the outcome?
External validity
extent to which experimental findings (specifically: causal
effects) may be generalized to other settings, measurements,
populations, and time periods
Between subject
– each subject is assigned to either treatment or control upon
entering the study
– comparison is made between groups of people (and require
assumption of equivalence between groups)
Within-subject
– each person serves as both treatment and control
– comparisons are made for a given person under treatment
and under control
Factorial designs
An experiment in which two or more variables (factors) are
manipulated independently
Natural experiments
experiments in which the intervention is
not under the control of the researcher
Payoffs of Natural Experiments
much stronger causal identification than traditional
case studies or correlational (quantitative) analyses
Types of Natural Experiments
- Randomizing device (with a known
probability) divides a population - Jurisdictional studies
- make use of geographic divisions to study similar populations
that find themselves by chances on opposite sides of some
divide - Omnibus (“other”) category
- e.g., effect of bad weather on economics
Are Large-N or small-N better for controlling for confounding variables?
Large-N
Degrees of Freedom
df : # observations − (number of IVs and controls)
Benefits of process tracing
→ addresses causal mechanisms
→ addresses concerns about reverse causation
Four types of process-tracing tests
- Straw-in-the-Wind
- Hoop tests
- Smoking Gun test
- Doubly-decisive
Straw-in-the-Wind Test
Can increase plausibility of a
hypothesis (or raise doubts), but are
not decisive by themselves
Hoop Test
hypothesis must “jump through the
hoop” in order to remain under
consideration
Smoking Gun Test
provides sufficient criterion for
accepting explanation, but is
not necessary
Doubly Decisive Test
provides necessary criteria and
sufficient criteria for accepting
explanation (confirms one hypothesis while ruling out all others)
motivational biases
we not only see what we expect to see, but particularly what
we want to see
Complexity
in a system (units/elements are interconnected), chains of
consequence extend over time and many areas
Fraud
falsifying data
Spurious correlation:
a correlation that is not what it
appears to be
P-hacking
the practice of reanalyzing data in many
different ways and only presenting the preferred results
The garden of forking paths
The unconscious tendency of individuals to fit their
processing of information to conclusions that suit some end or goal