Surveys

1
Surveys Questionnaires

• Data Collection

2
Data Will Be Collected On

• Sample
• Population

3
Sampling

• To conserve resources
• We use STATISTICS
• From REPRESENTATIVE SAMPLES
• To make INFERENCES
• About POPULATION PARAMETERS

4
Sampling The Jargon

• Population (universe) the total group or
  population being studied. All cases (units) that
  could possibly be studied.
• Sample a subset (a part of) the population

5
Sampling Implications

• Sampling involves selecting some cases for
  inclusion and many cases for exclusion
• Thus, there is always a chance that our
  conclusions about the population might be
  inaccurate, because of the samples
  characteristics
• We can reduce this chance to an acceptable
  minimum, but we cannot eliminate it
• More on this topic in a few weeks

6
Data Collection Method

• Experiments (already covered)
• Available data (already covered, under secondary
  sources)
• Self-reporting (surveys, polls, questionnaires)
• Observation
• Unobtrusive measures (physical evidence)
• Content Analysis

7
Self-Reporting

• Surveys (Interviews)
• Questionnaires

8
Self-Reporting Common (To Both Methods) Issues

• Length not too long, but long enough
• Open v. closed-ended questions
• Order of questions put the toughies towards
  the end
• Question repetition
• Wording
• Scoring

9
Wording Many Potential Problems

• Leading (bias)
• Two questions in one (double-barreled)
• Assuming too much knowledge
• Vagueness
• Contradictory, factual errors, etc.

10
Scoring

• SA A N/N D SD
• 1 2 3 4 5
• U.S. military forces are used too much as a
  primary instrument of American foreign policy.
• The U.S. has an obligation, as the dominant
  military power, to use its armed forces to
  promote legitimate goals.

11
Concepts, Operational Definitions Wording

• Concept Support for U.S. military involvement in
  overseas conflicts
• Operationalization Response on the previous
  slides Likert items

12
But

• Agreeing denotes support on one item
• Disagreeing denotes support on the other

13
A Note On Scoring

• Varying the wording so that the presence of the
  attribute being measured is indicated by
  agreement in some cases by disagreement in
  others is OK, even desirable, but
• Be aware of the scoring implications of your
  wording!

14
Self-Reporting Surveys (Interviews )

• What administered by trained interviewers
• Some issues
• Costs (face-to-face v. telephone)
• Quality of interviewers
• Training
• Supervision oversight

15
Self - Reporting Questionnaires

• What self-administered
• Some issues
• Response rate
• Respondents literacy (KISS!)

16
(No Transcript)