4.
Theories and
Theory Testing
January 21, 2021
Class Outline
I. Wrapping Up Political Concepts
II. Introduction to Social Science
Theory
1. What’s a Theory?
2. Types of Social Science Theories
3. Testing Theories OR What’s
Endogeneity?
4. Endogeneity in Practice
I.1 So Far: Documenting
“What”
I.2 Example: Measuring
Sustainability
I.3 One Response
1. Choose one systematized concept of sustainability to focus on:
Environmental Sustainability or Human Sustainability.
• Human Sustainability
2. Based on your choice, what indicators of sustainability would you
measure? What is your unit of analysis?
• My unit of analysis is the country. I will collect indicators of maternal
mortality as an indicator of sustainable development of a country’s
population over time.
3. Can you think of a “diminished subtype” of sustainability? In other
words a type that may possess many attributes of sustainability but
totally lack some dimension that makes it fail to qualify as sustainable?
• What if maternal mortality is very uneven in a country? For example, if one
ethnic group has massive maternal mortality problems, but it is a small
group, so the country’s data overall is close to average. Could this be
considered a diminished subtype of “uneven sustainability”?
II.1 From What to Why
• In other words, from concepts to theories
• From “What is democracy” to ???
• From “What is sustainability” to ???
II.1 What’s a Theory?
• Van Evera: “General
statements that describe
and explain the causes or
effects of classes of
phenomena.”
• SSRM: “A system of
constructs and
propositions that
collectively presents a
logical, systematic and
coherent explanation of a
phenomenon of interest.”
II.1 Some Related Terms
• Hypothesis
• Van Evera: “A conjectured relationship between two phenomena.
• Explanation
• Van Evera: “The causal laws or hypotheses that connect the cause
to the phenomenon being caused.”
• Boundary Condition
• Van Evera: “A phenomenon whose presence activates or magnifies
the action of a causal law or hypothesis.”
II.1 What’s in a Theory?
(SSRM)
• Concepts and Variables
• Independent and Dependent Variables
• Logic
• Assumptions/Boundary Conditions
II.1 An Example: Democracy
• Our Research Question: What causes some countries to be
democratic?
• Two Theories:
1. A strong middle class causes democracy, because the growth of a
middle class empowers individuals to want a say in their political
lives and gives them the means to organize for it.
2. A population with shared civic values causes democracy, because
shared values of tolerance and liberty make even losers likely to
respect the results of elections and winners likely to restrain their
use of power once in office.
• What are the independent variables and causal explanations in
these theories?
II.1 Group Example
• In groups of 3-4, answer the following questions related to the
following research question:
• Why do candidates win elections?
1. Identify two independent variables that could cause
candidates to win primary elections. State the hypothesis
relating each independent variable to the dependent
variable.
2. What is the explanation that ties each independent variable
to an election outcome?
3. Would you expect that there are any boundary conditions
on your theory? In other words, should your theory only
operate in certain contexts?
II.1 A Hypothesis, in Equation
Form
The
The Constant
Constant
Error
Error Term
Term
Dependent
Dependent Variable
Variable Independent
Independent Variable
Variable
Beta:
Beta: The
The Effect
Effect of
of the
the IV
IV on
on
the
the DV
DV
II.2 Varieties of Theories
• Where do theories come from?
• Inductive v. Deductive approaches.
• How far do theories extend?
• Nomothetic v. Idiographic
• “Specific explanations” in Van Evera.
II.2 What Makes a Good
Theory?
• Some Possible Standards:
1. Accuracy?
• How well does the theory fit a particular case?
2. Generalizability?
• How many cases can the theory help to explain?
3. Parsimony?
• How much explained with how little?
4. Falsifiability?
• Is it clear what evidence would refute the theory?
II.2 The Reality: Tradeoffs
II.3 How to Test Theories?
• A General Goal: Causal Inference
• Definition: The process of drawing a conclusion about a causal
connection based on the occurrence of an effect.
• Challenge: How do we really know if something caused an
event to occur?
• Relationship to “counterfactual” logic.
II.3 Methods of Testing
Causation
• Experimental Methods: The Researcher controls exposure to
the Independent Variable
• Usually statistical research
• Observational Methods: The Researcher observes the relevant
variables.
• Statistical, or Large-N, Research
• Qualitative, or Small-N, Research
II.3 A First Test: Correlation
II.3 A (Baked) Good Example
II.3 The Real Test: Exogeneity
• In your own words, what does it mean to say that “Correlation
does not equal causation.”?
II.3 The Importance of
Exogeneity
• Avoiding misleading theoretical conclusions.
• Essential to theory-testing and causal inference.
II.4 Exogeneity in Practice:
Country Music
II.4 Exogeneity in Practice:
Campaign Spending
For Next Time
• For SSRM/Bailey
• What is an experiment in political science? What makes a study
an experiment, specifically?
• What does randomization or random assignment mean in an
experimental research project, and why is it so important?
• Why are there some limits to the types of experiments we can do
in political science?
• For Eldersveld
• What basic research question is Eldersveld trying to answer?
• What makes this an experimental design? What is Eldersveld’s
“treatment” in the study?
