WORKSHOP: PART I

STATE & SOCIETY IN EAST ASIA

2021
What is a good research
question?

By the end of this workshop, you'll be able to:

1. Identify causal relations between variables;
2. Develop specific research questions from a topic;
3. Identify theories and methods as interconnected components of a
research project.
RESEARCH QUESTION

In comparative politics research, we don’t

investigate how the world should be.

We want to know why outcomes in the world

are as they have turned out.

We describe and explain via comparison, and

sometimes predict outcomes based on evidence
that we observe and collect.
SOCIAL SCIENCE

Social science rarely results in “proofs”

Most social science findings have exceptions

Relationships between variables are about probability and

likelihood, not “laws”

Unlike the physical and natural sciences, the social sciences often
cannot make use of controlled experiments (= test the hypothesis).
THEORIES
Non-normative research
Empirical verification /
testability
Falsifiable (Karl Popper)
Generalizations
The simplest explanation that fits
the evidence (aka parsimony
principle)
PHYSICS ENVY

To do “science” you need:

• To devise a theoretical model;
• To deduce a testable hypothesis from the model;
• To test the hypothesis against the world.
If you don’t do this, you are not a real
scientist… 
PHYSICS ENVY

To sound more “scientific” some political But very often political scientists can’t test their
scientists use evocative of physics and hypothesis.
chemistry: “theory,” “law,” or “experiment”…
 Concepts:

• Ideas comparativists use to think about the processes we

study
• Mental constructs which are a combination of
characteristics

Good concepts are:

• Clear and coherent

• Consistent
• Useful for measuring variables

CONCEPTS
 Conceptualization

• Identifying or making up concepts

• A process of defining the agreed meaning of the terms used in
a study

Operationalization

• Process of making basic concepts measurable

• Specific, concrete method to measure/manipulate a construct
• Measurables become “variables”
• Variables = operationalized constructs

DEVELOPING CONCEPTS
 CONCEPTS
Operational definition = a “recipe” for measuring
and/or manipulating a construct in a study.
WHAT ARE THREE CORE CONCEPTS FROM
THE FOLLOWING RESEARCH QUESTION
THAT COULD BE TURNED INTO
KEYWORDS FOR A SEARCH?
 How are gender roles portrayed in contemporary Japanese cinema?

 Why was the CCP able to preserve political stability during the market
liberalization in the 1980s?
Do you love
sports?
EXAMPLE # 1
Love of sports
Concept #1: the % of
life spent doing
sports
Concept #2: the % of
life spent watching
sports
 EXAMPLE #2

Develop an operational definition of

democracy.

What is democracy?
 EXAMPLE #2: DEMOCRACY
There are many possible operational definitions of
democracy.

For example:
 a country holds a free and fair multiparty election
 constitutional law guarantees freedom of speech, press, assembly, and religion
 there is no verifiable suppression of political participation and expression
 EXAMPLE #2: DEMOCRACY
Example: democracy and democratization
Clear and coherent
 Democratization cannot be vague: “all good things that happen when
a society changes”
 Can be “process by which rights and liberties are extended to all
adults in a country”
Consistent
 If democracy is defined as “a set of institutions” (not “values”),
democratization should also be about changes in institutions
Useful for measuring variables
 Ex.: democracy can be defined so that some places have it and some
do not
 EXAMPLE #3: DEVELOPMENT

Develop an operational definition of

development.

What is development?
 EXAMPLE #3: DEVELOPMENT
Development:
 Social development
 Economic development
 Political development
 Sustainability

Do we measure development in GDP growth?

Purchasing power?
Human right? Freedom of press? Education level?
Environment protection record? CO2 emission?
WHAT ARE THE KEY CONCEPTS?
OPERATIONALIZE THEM.
What external and internal factors played into South Korean and
Japanese development after WW II?
How does BRI infrastructure projects influence local socioeconomic
development?
 CASES & CASE STUDIES
Cases:
 units of analysis in social science
 examples of phenomenon to be studied

What could be a case? How to find a good case?

What is a case study?

HOW MANY
CASES IS
ENOUGH?

The study of a single case is considered comparative if it uses or

develops concepts applicable to other cases, and/or seeks to
make larger inferences.
Ideal to examine a deviant case, to generate hypotheses, to
develop new classifications.
Inferences based upon one case are less secure.
7 TYPES OF CASES
(GERRING & SEAWRIGHT 2008)
Typical (cases that confirm a given theory)

Diverse (cases that illuminate the full range of variation on X, Y or X/Y)

Extreme (cases with an extremely unusual values on X or Y)

Deviant (cases that deviate from an established cross-case population)

Influential (cases with established and influential configurations of X’s)

Most similar (cases are similar on all variables except X1 and Y)

Most different (cases are different on all variables except X1 and Y)

DEDUCTIVE REASONING
Move from general theories to
specific observations or predictions
about a set of cases.

Theory Hypothesis Observation Argument

(general) (specific)
INDUCTIVE REASONING

Observation Pattern Hypothesis Theory

(general)
 Variable:
• A principal element or factor that can
change, or vary, from case to case
• Any measured characteristic or attribute
that has the potential to differ from case
to case

VARIABLES Focus on causes and effects

Effects also called “consequences”

or “outcomes”

X independent variable → Y
dependent variable
INDEPENDENT VS. DEPENDENT VARIABLES
Positive feedback & self confidence

Headache & aspirin

Democracy & economic growth?

Media & public policy?

 Intervening variable explains relations

INTERVENING between X and Y.

VARIABLE Ex.: Poverty (X) → Lack of access to

healthcare → Short life (Y)
THE COMPARATIVE METHOD:
VARIABLES AND COMPARISON
Example: two country cases, Ghana and Togo
The two cases have different “regime types”
 Ghana is democracy
 Togo is authoritarian

Why the variation between the two cases?

 We want to explain difference in the variable “regime type”
(democracy vs. authoritarian)
 In our comparative question, regime type is the outcome (or Y
dependent variable)
 We look for the causes (or X independent variables) in the two cases
that explain the variation in regime type
THE COMPARATIVE METHOD:
BETWEEN-CASE COMPARISON

Comparative analysis of two or more cases

 Most Similar Systems design
 Most Different Systems design
 THE COMPARATIVE METHOD
Most Similar Systems Design: Research design in which we
compare cases that are similar with respect to a number of
factors, but with distinct outcomes
Two (or more) cases similar on many variables
 But obtain different outcomes
 Logic is that we can rule out (or “control for”) many possible
variables that are similar
 Leaves a few other variables to explain outcomes
THE COMPARATIVE METHOD:
THE MOST SIMILAR SYSTEMS DESIGN
WHY DO THESE CONTIGUOUS COUNTRIES IN AFRICA HAVE DIFFERENT
REGIME TYPES?

Variables Case 1: Togo Case 2: Ghana

Similarities
Climate Hot / tropical Hot / tropical
Income Low Low
Ethnic demography Heterogeneous Heterogeneous
Largest religion Christianity Christianity
Other religions Islam, traditional Islam, traditional
Outcomes
Regime Type Authoritarian Democracy
Cause
Hypothesis: Colonizer France United Kingdom
 THE COMPARATIVE METHOD
Most Different Systems Design: Research design in which
we compare cases that differ with respect to multiple factors
but in which the outcome is the same
Two (or more) cases differ on many variables
 But end up with the similar outcomes
 Logic is that we can rule out (or “control for”) many possible
variables that are different
 Leaves a few other variables to explain outcomes
THE COMPARATIVE METHOD:
THE MOST DIFFERENT SYSTEMS DESIGN
WHY DO MAJOR SOCIAL REVOLUTIONS OCCUR?

Variables Case 1: Case 2:

France 1780s China 1940s
Differences
Continent Europe Asia
Population < 30 million > 500 million
Century 18th 20th
Regime Monarchy Nationalist party
Outcomes
Social Revolution YES YES
Cause
Hypothesis ??? ???
THE COMPARATIVE METHOD
Within-Case Comparison: Comparative analysis of
variation over time or in distinct parts of a single case

Digs into a case to better understand evidence

Observational / relational questions are designed to look at the relationships
between two or more variables and provide a general description of phenomena or
events.

How did this happen?

• How did South Korea democratize?

• How does national identity develop among a population that has troublesome political borders?

What are the key features?

• What are the key features of the South Korean developmental state?
• What are the major characteristics of the China Model?

OBSERVATIONAL QUESTIONS
Causal Questions: designed to determine whether one or more
variables cause or affect one or more outcome variables.

Contrast with “who, what, where, when” questions

• Ex.: In 1492, Columbus sailed the ocean blue
• Tells who, what, where, when
• “Why” is more difficult to answer (and “Why” is often debatable)
• Answering “Why” requires theories and evidence

These are questions of cause and effect

• “Why” does something happen?
• “Because…” (note the root word: cause)

CAUSAL QUESTIONS
CAUSAL QUESTIONS

Causal questions lead to empirical arguments

• Drawn from observations of facts in the world
• Explanations about why the world is the way it is
This is distinct from normative questions and moral
arguments
• Concerned with what is moral or ethical
• Arguments about how the world ought to be

Comparative study = puzzle = investigation

 CAUSAL QUESTIONS
It is possible to ask questions that don’t begin with “Why”
 What are the consequences of different kinds of institutions for
policy?
 What are the consequences of presidential vs. parliamentary
systems for education policy?
 Under what conditions will democracies emerge and consolidate?
 How do major social revolutions affect subsequent political
developments in given countries?
These questions are about cause and effect
HYPOTHESIS TESTING:
CORRELATION AND CAUSATION

Correlation
• Relationship between two variables in which they tend to move in
a predictable relationship to one another
• Can be in same direction (positive correlation)
 Ex.: > wealth is related to > democracy
• Can be in opposite directions (negative correlation)
 Ex.: > amount of absolute poverty is related to < democracy

 Can be false: an increase in ice cream sales correlates with

increased crime

Causation
• A Relationship in which one thing causes another
• In comparative politics, we try to understand causation between
variables.
CORRELATION AND CAUSATION
X variable → Y variable
 Variable Y may be caused by many X variables

Cause (X) → Effect (Y) X Y

Independent variable (X) → Dependent

variable (Y) EXAMPLE: The greater the cleavage
between governing elites in an
autocracy, the greater the
Explanatory variable (X) → Outcome movement towards a regime
(Y) transition.
 CORRELATION AND CAUSATION
PROBLEM: DEFINITIONAL PROBLEMS OR FALSIFIABILITY
PROBLEM

X and Y will correlate if

they are the same thing X = Y
Ex.: Dictatorship occurs
because of actions of the
dictator
 This is circular or
tautological argument
CORRELATION AND CAUSATION
PROBLEM: REVERSE CAUSATION
Y causes X vs. X causes Y
Argument that X causes Y has
causality reversed
Ex.: >cancer rates leads to
>longevity by country
X Y
-High cancer rates do not cause
longevity
-Longevity can lead to higher rates of
cancer (affects many older people)

Legitimacy of the CCP & Income rise in

China
Legitimacy → Income rise?
Income rise → Legitimacy?
CORRELATION AND CAUSATION
POTENTIAL PROBLEM: ENDOGENEITY

X causes Y, but Y also causes X

Ex.: health and education
-Better health leads to better
education X Y
-Better education leads to better
health
An increase in modernization leads
to increase in civil society AND
Not always a “problem” as long as
analyst recognizes it and does not An increase in civil society leads to
make wrong claims. an increase in modernization.
CORRELATION AND CAUSATION
POTENTIAL PROBLEM: INTERVENING VARIABLE
X leads to Y, but only indirectly (say, via
variable Z)
Ex.: Communism and democratization in
Eastern Europe
-Communism leads to protests or reactions
against tyranny Z
-These reactions culminate in democratization
-Erroneous to say communism led directly to
democracy

X Y
Intervening variables are not a “problem”
(and can be part of an answer) as long as
they are included in analysis.
WHY ONLY CRIMEA?
Transnistria (Russia VS Moldavia), 2006
South Ossetia and Abkhazia (Russia VS Georgia), 2008
Lugansk and Donbas (Russia VS Ukraine), 2014-2015
Crimea (Russia VS Ukraine), 2014

Russian-speaking population?
Western rivalry?
Conflict / Civic war?
Imperial revival?
Vladimir Putin?
CORRELATION AND CAUSATION
PROBLEM: OMITTED VARIABLE

X correlates with Y, because both are

caused by another variable Z.

Ex.: Ice cream sales & Murder rates

There is a positive correlations between Z
ice cream sales and murder rates: both
increase monthly.
Eating ice cream → murder?
Murder → buy an ice cream?
X Y
No. An omitted variable is weather.

This is a quite common error in social

science.
CORRELATION AND CAUSATION
PROBLEM: SPURIOUS CORRELATION

X and Y correlate, but no good logic

explains why
 If X and Y were to have causal
relationship, they would also correlate
X ? Y
with other factors.
Observational / relational questions are designed to look at the relationships
between two or more variables and provide a general description of phenomena or
events.

How did this happen?

• How did South Korea democratize?

• How does national identity develop among a population that has troublesome political borders?

What are the key features?

• What are the key features of the South Korean developmental state?
• What are the major characteristics of the China Model?

OBSERVATIONAL QUESTIONS
CAUSAL QUESTIONS

Causal questions lead to empirical arguments

• Drawn from observations of facts in the world
• Explanations about why the world is the way it is
This is distinct from normative questions and moral
arguments
• Concerned with what is moral or ethical
• Arguments about how the world ought to be

Comparative study = puzzle = investigation

 GOOD RESEARCH QUESTIONS
1. Your question must be researchable using primary and/or secondary sources
GOOD: How does the politics of history and historical memory influence China-
Japan security relations?
BAD: What do Shinzo Abe and Xi Jinping think about World War II?
>>> To answer this question, you need to examine Abe's and Xi's personal
perceptions of World War II. You might do so only by doing interviews with
them and/or studying their memoirs.
 GOOD RESEARCH QUESTIONS
2. Your question must be researchable using primary and/or secondary sources
GOOD: How does the politics of history and historical memory influence China-
Japan security relations?
BAD: What do Shinzo Abe and Xi Jinping think about World War II?
>>> To answer this question, you need to examine Abe's and Xi's personal
perceptions of World War II. You might do so only by doing interviews with
them and/or studying their memoirs.
 GOOD RESEARCH QUESTIONS
3. Your research question should be specific enough to answer it thoroughly.
GOOD: How did attitudes towards China change in the United States after the
COVID-19 outbreak?
BAD: How did attitudes towards China change around the world after the
COVID-19 outbreak?
>>> This question is very vague. It will be difficult to cover "the world" in an
essay or a thesis.
 GOOD RESEARCH QUESTIONS
Always narrow down your questions. For example, your topic of interest is
democracy, and you are interested in the process of democratization (an
introduction of a democratic system or democratic principles in a country). In this
case, you could focus on the following question:
 Why are some countries democratic and others not?

However, you need to make this question specific by adding case(s). For
example:
 Why is Japan a democracy?
 Why is China not a democracy?
 Why is Japan a democracy and China not?
 How did Japan's relations with the United States impact its democratic transition after World War
II?
 GOOD RESEARCH QUESTIONS
4. A balanced research question is specific yet complex enough to develop the
answer over the space of your thesis
GOOD: How and with effect did Japan government change after World War
II?
BAD: What are the four ways the Japanese government changed with the
MacArthur Constitution?
>>> This question is too simple. It can be looked up online and answered in a
few descriptive sentences. Thus, this question leaves no room for analysis.
 GOOD RESEARCH QUESTIONS
5. Your questions should also be relevant to your field of study and/or society
more broadly.
If you cannot explain why this question is worth asking, it is probably not your
best option.

6. Finally, ask yourself: is it feasible to answer this question within the timeframe
and practical constraints?
In other words, will you be able to finish your essay/ thesis on time? If your
research question is too big, you might want to narrow it down by reducing the
number of cases or limiting the timeline of your study.
EXERCISE
In Jan. 2021, a South Korean court ruled that the government of Japan
should pay compensation to 12 women forced to work as sex slaves for
Japanese soldiers during the second world war. However, Japan
maintains that several agreements have already settled all such claims.
Drawing on this event, formulate two causal research questions.

https://docs.google.com/forms/d/16iAmuki7TvQy2aOYnBzsgVeca9
w89-gmnjmLqmR9JN8/edit
JUST DON’T

Workshop: Part II
EVERYTIME YOU
CITE WIKIPEDIA
A KITTEN DIES

Please
don’t cite
Wikipedia.
IT IS VERY IMPORTANT TO NOTE, HOWEVER…
DO NOT USE CONTRACTIONS IN FORMAL WRITING.
OF COURSE!
 TABOO WORDS IN ACADEMIC
WRITING
Using vague terms makes your writing imprecise and may cause people
to interpret it in different ways. Always try to be as specific as
possible.
Taboo Example Alternative
Stuff People are concerned People are concerned
about their stuff about
their (belongings/possessi
ons/personal effects)
Thing The report presents The report presents
many things many (details/findings/rec
ommendations)
A long time, a while This topic has interested This topic has interested
researchers for a long researchers for more than
time 30 years
TABOO WORDS IN ACADEMIC
WRITING
Academic writing is usually unadorned and direct.
Avoid dramatic vocabulary:
 adverbs of frequency (such as always and never);
 superlatives (terms that indicate something is of the highest degree, such as the best);
 intensifiers (words that create emphasis, such as very)

May also not be accurate – are you 100% sure that something
is perfect or never happens?
These terms do sometimes add value, but try to use them sparingly.
Taboo Example Alternative

Always, never Researchers always argue Researchers (frequently/

that commonly /
typically) argue that

Perfect, best, worst, most The perfect solution to the (An ideal solution/one of
(or any other superlative) problem the best solutions) to the
problem

Very, extremely, really, This theory This theory is (important/

too, so (or any other is extremely important critical/ crucial)
intensifier)
Taboo Example Alternative

Beautiful, ugly, wonderful, A review of literature A review of the literature

horrible, good, bad yielded yielded
many good articles many relevant articles

Naturally The participants naturally The participants wanted to

wanted to know know

Obviously, of course The The

results obviously indicate results clearly indicate
TABOO WORDS IN ACADEMIC
WRITING
In general, you should also try to avoid using words and phrases that
fall into the following categories:
Jargon (i.e., “insider” terminology that may be difficult for readers
from other fields to understand)
Clichés (i.e., expressions that are heavily overused, such as think
outside of the box and at the end of the day)
Everyday abbreviations (e.g. photos, fridge, phone, info)
Slang (e.g. cops, cool)
Gender-biased language (e.g. firemen, mankind)
DON’T
PLAGIARIZE!