Path: > Syllabus > Outline > Topics and Readings > Quantitative Analysis > Statistics and the Logic of Inquiry

Quantitatitive Analysis in Political Research: Lecture 2
Statistics and the Logic of Inquiry
Lecture 2
  •  Statistics and theory
    • Two types of theory in political science: normative and empirical
    • Traditionally, "political theory" was concerned with normative matters -- "political philosophy"
      • 1.Examples of normative concepts: freedom, order, equality
      • 2.Examples of normative analysis: conflict between freedom and order, between freedom and equality
    • Statistical analysis is concerned with empirical theory.
      • Quantification in general forces one to become more rigorous in making statements
      • Rigorous theoretical thinking requires careful attention to form and content of statements.
      • Two types of statements
        • SINGULAR: about particular things--Clinton is a Democrat.
        • GENERALIZATIONS: Republicans are more conservative than Democrats.
      • Two types of generalizations: Deterministic (E=mc2) and probabilistic
    • Statistical analysis is concerned with testing the "truth value" of probabilistic generalizations
      • Statistics is best suited for analysis of similar social processes rather than unique events.
      • Theoretical assertions about these events must be explicit and unequivocal -- they must be FALSIFIABLE.
      • Science advances more readily through error than confusion.
    • Two types of statistics:
      • DESCRIPTIVE STATISTICS: describe and summarize data
        • Generalize beyond data at hand
        • Evaluate differences between groups
        • Estimate unknown values
      • Our course will be organized to take up descriptive statistics first, then the more complex inferential statistics
  • Statistical analysis should be tied closely to theory construction.
    • What is an empirical theory?
      • A theory is a set of interrelated propositions
        • A proposition is a statement of relationship between concepts
          • A concept is a general idea for grouping phenomena as similar
  • Example: politicoeconomic mini-theory (from Bohrnstedt and Knocke:Statistics for Social Data Analysis)
    • Propositions

      P1: Economic instability generates disaffection with the national political regime
      P2: Disaffection with the national regime strengthens the opposition political forces

    • Deduction:

      P3: Economic instability increases strength of political opposition

    • Theoretical terms
      • Scope conditions:
        • space
        • time
      • Units of analysis:
        • individuals,
        • spatial aggregates of individuals,
        • organizations
  • Testing a theory
    • Remember, the goal is to render it FALSIFIABLE
    • The abstract concepts in the proposition must be made concrete
    • Done through operationalization
      • The specified "operations" that must be performed to measure the concept
      • Often multiple indicators of complex concepts are desirable, but we will consider only single indicators here.
  • Example of the politicoeconomic mini-theory:
    • Economic instability --------> high inflation rates
    • disaffection with national regime ----> negative attitudes toward the president's economic policies
    • strengthens the opposition's political forces ----> increases support for party not in the White House
  • Adequacy of operationalizations
    • validity -- or accuracy
    • reliability
  • Causal terminology for concepts or variables in a theory
    • Independent variable -- the causal agent
    • Dependent variable -- the caused variable
  • Graph of the mini-theory
    • Propositional form:
    • Hypothesis: