Techniques of Political Analysis

1
Techniques of Political Analysis

• Introduction to Science
• The scientific method, research design and
  quantitative methods

2
Defining concepts

• CONCEPTUALIZATION
• Conceptual definition the more general
  definition of a concept such as one would find in
  a textbook or dictionary

3
Operationalization

• Operational definition a measurement definition
  defines how we are going to MEASURE the concept

4
Why Study Statistics?

• Its better to know some of the questions than
  all of the answers
• --James Thurber

5
Statistical Definitions

• Statistics-- the method of describing and
  interpreting data
• Data contain information about some group of
  individuals and is organized in variables
• Individuals are the objects described by a set of
  data.May be people, animals, or things
• A variable is any characteristic of an
  individual. Can take on different values for
  different individuals

6
Measurement

• The systematic observation and representation by
  scores or numerals, of the variables we have
  decided to investigate
• Provides bridge between our proposed explanations
  and the empirical world
• Quality of measurements is judged in terms of
  both their accuracy and precision

7
Accuracy

• Reliability-Concerns the extent to which an
  experiment, test, or any measuring procedure
  yields the same results on repeated trials. The
  more consistent the results given by repeated
  measurements, the higher the reliability of the
  measuring procedure
• ValidityThe measure should reflect the true
  value of what is being measured

8
Reliability
Measure should yield the same results on
repeated trials-- values should be consistent
High reliability
Low Reliability
9
Reliability

• Yardstick-- ordinary wooden vs. elastic material
• votes cast in an election
• Machine tally
• Hand recount
• Chads--Pregnant chads, hanging chads, dimpled
  chads

10
Validity
The measure should correspond to the concept it
Operationalizes. It measures what it purports to
measure
High Validity
11
Validity

• Validity is very important in polling since all
  polls assume the sample they draw is not
  systematically different from the population.
• Literary Digest--presidential polls
• 1924- owners of cars or telephones
• 20 response
• 1936-- predicted Landon

12
Error

• Valid measures are generally reliable, but
  reliable measures can certainly be invalid
• Reliability is affected by random error
• Validity is affected by systematic bias

13
Checks for Reliability

• Test-retest (stability)
• Alternative-form (equivalence)
• Split- half (homogeneity)
• Inter-coder

14
Test-retest

• Simply consists of repeating the measurement a
  second time, allowing for a suitable interval of
  time between the two measures
• Looking for temporal consistency
• Take SAT twice, ask survey questions twice
• Problems?

15
Alternative -form

• Also, measures concept more than once, but with
  two different measures.
• Measure how good a student is-- GPA, SAT,
  evaluations
• Helps with instrumentation influence, but time
  still an issue

16
Split- half

• Measures internal consistency
• Divide the measure in half to see if the whole
  thing taps the same concept
• Multi-item measures-- IQ test, political
  knowledge quiz

17
Inter-coder

• Two individuals should calculate the same values
  using the same coding scheme
• Florida votes
• Segal and Cover

18
Checks for Validity

• Face Validity
• Content Validity
• Construct Validity
• Interitem Association
• Criterion Validity

19
Face Validity

• The extent to which the measure is subjectively
  viewed by knowledgeable individuals as covering
  the concept-- the instrument APPEARS to measure
  the concept

20
Content Validity

• Does the measure covers all of the generally
  accepted meanings of the concept?
• Measure of liberalism by asking about womens
  rights

21
Construct Validity

• When a measure of a concept is related to a
  measure of another concept with which the
  original concept is thought to be related,
  construct validity is demonstrated.
• Education and income
• Political efficacy and political participation
• GRE example--

22
Interitem Association

• Multiple measures of the same concept correlate
  positively with one another-- correlation matrix
• Economic health-- inflation, unemployment,
  deficit

23
Criterion Validity

• Determines whether our measure is able to predict
  some criterion external to it.
• Aptitude tests and other predictive tests

24
Increase Reliability and Validity

• Use multiple measures
• Index--
• Scale

25
Precision

• Measurements should contain as much information
  as possible about the attribute or behavior being
  measured
• Precision is reflected in the level of
  measurement of the variable

26

• Game day weather is related to how well the OSU
  football team does
• Could be dichotomous-- win lose
• Could be ranking (over 40, from 20-39, under 20)
• Could be actual points scores
• Which one is best, provides most information,
  best operationalizes the concept?

27
Levels of measurement

• Nominal
• Ordinal
• Interval
• Ratio

28
Nominal

• Nominal values assigned to a variable represent
  only different categories or classifications--
  exhaustive and mutually exclusive
• Represents attributes not quantities
• Campus Buildings
• Religion

29
Ordinal

• Ordinal assumes that more or less of a variable
  can be measured and that a comparison can be made
  on which observations have more or less of a
  particular attribute-- does not include the
  magnitude of differences between numbers
• Line up by height, teacher evaluation scales

30
Interval

• Interval The intervals between the categories or
  values assigned to the particular observations
  have meaning.For interval measures, the value of
  a particular observation is important not just in
  terms of whether it is larger or smaller, but HOW
  MUCH larger or smaller
• Temperature, years

31
Ratio

• RatioThe values of the categories order the
  categories, tell something about the intervals
  between the categories, and state precisely the
  relative amounts of the variable that the
  categories represent.
• Value of zero actually represents the absence of
  that variable
• Age, money, votes