Please enable JavaScript.
Coggle requires JavaScript to display documents.
Comparative Politics - Coggle Diagram
Comparative Politics
Developing a research design
research questions are theory guided
relationship between dependent variable (Y) and independent variable (X)
research answers are interpreted from evidence
research design links research answer and research question
validity of research design
internal validity
how accurate your conclusion is
external validity
whether your results can apply to units not studied
types of multi-method research
Combining qualitative approaches
eg using case studies and interviews
Combining qualitative and quantitative approaches
eg statistical analysis and process tracking
Process tracking
qualitative method
complex causal processes by means of historical narratives are evaluated
focuses on causal mechanisms
how X produces the conditions that produce Y
used in hypothesis testing
Combining conventional qualitative methods with constructivist or interpretivist approaches
Case selection
Single-case study
used as a pilot for generating theories
can invalidate extant theories
Time series
used to analyse which factors have become more relevant over time
Closed universe
comparing a few cases at different points in time
taking into account defining period intervals
eg analysing inter-war period
Cross section
several cases compared simultaneously
most commonly used
Pooled analysis
pooling cases across time and systems
very advanced analysis
What is comparative politics
involves using many different skills to compare units
inference
does one thing cause another
description
who votes for what, where and when
are people happy with the state/government
explanation
what produces differences
prediction
anticipate outcomes based on observations
sharpen definitions and improve concepts
What is compared
units
countries/regions
elections
eg are particular events causing particular results
parties
individual voters
why would voters vote against their interests
Objects
institutions/actors
outcomes
policy/votes/stability
dimensions
across space and time
how the past shapes the present
across levels
individual
group
national
across issue areas
Different approaches
different approaches highlight different causes
institutions
electoral systems
how votes are counted
constituencies
majoritarian or proportional system
party competition
how many parties are there
do the parties represent all the people
media rules
regulation
communication between government and public
Interests
economic insecurity
voters mostly care about the economy
welfare expectations
Ideas
Beliefs/values
national identity
immigration attitudes
prejudices
international factors
cross border shocks
war on drugs
migration
crises
individuals
psychology
biases
motivated reasoning
if they think they are right, they refuse to listen to other viewpoints
Identity-protective cognition
wants to protect national identity
emotions
anger
different lenses
positivism
values distinction and implying that there are real facts
observable and verifiable results
can be studied the same way natural sciences are
Constructivism
facts are social constructions
researchers can't be objective
Marxism
class conflict explains the difference among political systems
Institutionalism
structures shape politics
new institutionalism
individuals act on the basis of their connections
4 alternative conceptions of institutions
Normative institutionalism
institutions composed of norms that shape behaviour
Rational choice institutionalism
individuals are self interested
utilise incentives of the institution
Historical institutionalism
role of ideas
Empirical institutionalism
do institutions even matter
Methodological individualism
individual choices rather than institutional constraints produced observed differences in governments
Conducting comparative research
variables
research question
Y
What are you trying to explain
X
whom/when/where
concepts
define them
choose indicators - eg what data will you use
ensure comparability
pick a good selection of cases/units
methods of study
Large-N studies
good for patterns and generalisations
large number of cases but less detail
Small-N studies
good for mechanisms and context
small number of cases in more detail
Experiments
micro level
good for narrow setting and causal identification
mixed method
Use large-N to find patterns
use small-N to explain why
good for creating robust stories
Eg - chance of populists taking power
Y - probability of voting for a populist party
X1 - economic security
X2 - prejudice, immigration attitudes
Controls - age, education, region
can see if same factors are present in different areas and this can be used to infer future events
Two classic comparative designs
most similar systems design
Pick similar contexts
look for key differences
Eg - nordic countries with similar welfare states but different populist vote trajectories
hypothesis focus
institutional rules
party strategies
media environment
most different systems design
Pick very different contexts
look for shared factors
Eg - do populists succeed in both old and new democracies
Hypothesis focus
common psychological mechanisms