Causality

1
Causality Research DesignPart 1

• October 6, 2004
• Sociology 487

2
Lecture Topics

• Nomothetic and Idiographic Causal Explanations
• Types of Causes
• Criteria for Determining Causality

3
Idiographic Causal Explanations

• Individual sequence of events that result in a
  particular outcome for a particular individual
• Emphasis on context
• Holistic
• Deterministic

4
Nomothetic Causal Explanations

• Change in independent variable leads to change in
  dependent variable, ceteris paribus
• Ceteris Paribus means all else equal
• Probabilistic
• The counterfactual
• Situation as it would have been in absence of
  variation in independent variable

5
Types of Causes

• Necessary cause
• Condition that must be present for outcome to
  result
• Sufficient cause
• Condition that, if present, ensures outcome

6
Major Criteria for Determining Causality

• Temporal priority of the independent variable
• Empirical association
• Nonspurious relationship

7
Temporal Priority
The change in X must occur before the change in Y.
?X
?Y
t1 t2
8
Empirical Association
The independent variable and the dependent
variable must vary together.
A change in X is associated with a change in Y.
?X
?Y
9
Spurious Relationship

• Relationship that is explained away by variables
  other than those included in the research
  hypothesis

10
Reducing Risks of Spurious Relationship

• Random assignment
• Statistical control

11
Two Additional Criteria for Determining
Causality

• Identifying a causal mechanism
• Specifying the context in which the effect occurs

12
Identifying Causal Mechanisms

• Identifying process through which X affects Y
• Intervening variables

13
Intervening and Extraneous Variables

• Intervening Variables
• Helps explain the effect of X on Y

• Extraneous Variables
• Explains away the spurious relationship between X
  and Y

14
Determining Context in which Effect Occurs

• When does effect occur?
• For whom does effect occur?
• Under what conditions does effect occur?

15
Causality and Research Design

• Design of your study will help you assess
  causality
• Cross-sectional versus longitudinal designs
• Use of different units of analysis

16
Key Points

• Both idiographic and nomothetic explanations are
  concerned with causation
• Three basic criteria for establishing nomothetic
  causation
• X must precede Y
• X must be associated with Y
• Effect of X on Y is not due to effect of Z on
  both X and Y
• Two additional criteria used to assess causality
  
• Specification of causal mechanism through which X
  affects Y
• Specification of the context in which X affects Y