Chapter 11: Survey Research Summary page 343

1
Chapter 11 Survey ResearchSummary page 343

• Asking Questions
• Obtaining Answers
• Multi-item Scales
• Response Biases
• Questionnaire Design
• Questionnaire Administration
• Survey Data Archives

2
Purposes and Types

• Two purposeswhat are they?
• Estimate population parameters
• Test hypotheses
• ? Which type can be tested? Corr or Experimental?
• Both! But how would you conduct an experiment?
• Two types of recruiting Ps are?
• Sample surveys
• Convenience surveys

3
Before Asking Questions

• Establish purpose(s)
• The purpose of this survey is to determine
• Specifics .e.g. Ps attitudes toward, intentions,
  etc.
• Characteristics of Ps (attributes)
• Beliefs (cognitive)
• Feelings (affect)
• Behavioral intentions
• Past behaviors
• Objectives (Explicit)
• Specify what findings will indicate the
  objectives were met
• Anticipate what analyses will need to be
  conducted (ahead of time!)

4
Before Asking Questions

• Determine sample demographics needed
• Grouping variables
• E.g. race, gender, age, occupation, dept,
  geographical
• Determine constructs
• Established measures
• E.g. job stress
• Self-developed measures
• E.g. Attitudes toward drug therapy
• Single item variables
• E.g. intention to continue or quit
• Determine topic and question order

5
Asking Questions Open v. Closed Ended

• Open Qs when
• For some sensitive Qs
• When dont know all options
• Esp. exploratory study (use interview/focus
  group)
• Certain types of judgments
• Caution
• Open ones have to use content analysis
• ? What are the relative advantages and
  disadvantages of open and closed-ended questions?

6
Asking Question Wording

• Use Ps language
• To hell with grammatically correct
• Clarity takes precedence
• Avoid Negatives (double ones)
• But counterbalance to avoid acquiescence
• One at a time
• No double barreled
• would you like to pass this course and get an A?

7
Asking Question Wording (cont)

• No leading questions
• Dont you agree that Dr. Mitchell is an
  excellent professor?
• Be specific
• Do you like your instructors here at UB?
• Dont make assumptions
• How long have you been a thug?
• Address sensitive topics sensitively
• Even the best students get irritated with
  professors sometimes

8
Obtaining Answers

• Levels of measurement
• Consider the construct
• Consider precision needed
• ? What are things to consider in determining what
  level of measurement to use for a question?
• Information content
• Statistical tests
• Decide ahead of time!
• Ecological Validity
• E.g. mock jury (interval) v. real jury (nominal)

9
Response Format p 354

• Choice depends on purpose
• E.g. should each of a category be itemized?
• Comparative rating scales
• Method of paired comparisons (all possible
  combos)
• Itemized rating scales (MC questions)
• Make sure responses are mutually exclusive
• Nominal, ordinal
• Graphic scales
• Visual anchors
• Numerical rating scales
• 5,7,9 points with labels (interval level)

10
Response Formats cont

• Multi-item scales
• Likert scales
• Consider direction reverse some or not
• Thurstone
• Not worth the trouble
• Guttman scales
• Ordered attitude items with increasing levels of
  agreement
• marry 1. stranger, 2. cousin, 3. sibling. 4. ?
• Semantic Differential
• Choose relevant adjectives

11
Response Biases

• Question-related
• Scale ambiguity (choose meaningful labels)
• Category Anchoring (find right range)
• Estimation biases (open ended)
• Retrospective is very unreliable
• E.g. self-reported illegal drug use
• Respondent interpretations of numerical scales
• E.g. not at all successful?

12
Response Biases

• Person-related
• Social desirability
• Acquiescence
• Extremity response rates
• (gender, age, race differences)
• Halo Leniency
• Interpreting responses
• Literal interpretation fallacy
• A good reason to calculate means

13
Questionnaire Design

• Question order
• Question sequencing
• Easy to difficult related to purpose
• General to specific within topic
• Context effects
• An early question may influence response to a
  later one

14
Questionnaire Design

• Layout
• Use closed as much as possible
• Consistent format
• depending upon the nature of constructs
• Whether or not you want to compare to norms
• Instructions
• Used to motivate and clarity how to respond
• Use existing measures
• Response formats

15
Methods of Data Collection

• Group (efficient)
• Mail (low response rate)
• Personal (good but expensive)
• Phone (efficient and provide anonymity)
• Focus groups
• Confounding, conformity, all sorts of problems
• Computer (lots of advantages)
• But identity confirmation, duplicate entry
  problems
• Compare methods (use table 11-12, p. 382)

16
Survey Data Archives

• Great, but never seem to answer your research
  questions
• But start there
• ? Formulate a hypothesis you could test with a
  survey.

17
Survey Research summary

• Asking Questions
• Obtaining Answers
• Multi-item Scales
• Response Biases
• Questionnaire Design
• Questionnaire Administration
• Survey Data Archives