Behavioral Attitude Measurement Theory

1
Behavioral Attitude Measurement Theory

• Don W. Stacks, Ph.D.
• School of Communication
• University of Miami
• Coral Gables, FL 33124
• and
• The Institute for Public Relations

2
What is Measurement?

• Assigning numbers to objects or concepts
  systematically
• With ways to assess measurement validity and
  reliability

3
Assessing Objectives via Measurement

• A way of assessing public relations objectives
• Simple counts informational
• Complex thematic Motivational
• Observational counts Behavioral
• Via data
• Categorical (nominal, ordinal , )
• Continuous (interval, ratio Mean, Median, Mode,
  variance)

4
Measuring Behavioral Objectives

• Opinions ? Behaviorbut consist of verbal
  expressions of
• Attitudes
• Beliefs
• Values

5
Attitude Dimensions

• Cognitive (informative)
• Affective (emotional)
• Connotative (behavioral)

6
Attitude Scaling

• Validity
• Reflects the objective of the measurement
• Can be assessed qualitatively and statistically
• Reliability
• Reflects stability in measurement
• Can be assessed statistically

7
Creating Attitude Scales

• Usually asks Ps to respond to an attitude
  statement on a continuum.
• Response is closed-ended
• Measurement is good SA A N D SD
• This panels participantsdid a good job SA A
  N D SD

8
Requirements (contd.)

• Must allow for a neutral response
• To be valid each item must contain an odd number
  of responses (3, 5, 7, 9)
• SA A N D
  SD

9
Requirements (contd.)

• Responses must be balanced
• Excellent antonym is Terrible, not Bad
• Excellent
• Bad
• Must consist of of least 2 or more statements
• (to assess for measurement reliability)

10
Things to Avoid

• Double-barreled statements
• Watch statements with and, or or
• Statements with an even number of responses
• SA A D SD (No response ?
  neutral)
• Single-item statements
• Unbalanced responses
• Excellent Good Neutral Poor Bad

11
Reliability Validity
unreliable
Reliable but not valid
Reliable and valid
12
Relationship Between Measurement Reliability
Validity

• A measure must be reliable or you cannot assess
  its validity!