VII. SCALES AND INDEXES

1
VII. SCALES
AND INDEXES
2
A. Logic of Scaling

• 1. the practical application of
  conceptualization
• 2. given complex issues, multiple
  measures leading to a
• single conclusion is often required
• 3. rationale
• a. need for multiple indicators
• b. statistical analysis requires
  measurement variation
• c. given complex attitudinal
  measures, tendencies and
• patterns become more easily
  discerned through
• multiple indicators

3

• 4. definitions
• a. scales composite, multiple
  measures about a
• particular aspect
  of a theoretical concept
• 1) scales are used to measure
  your theoretical
• concerns
• 2) these are generally
  measures of your
• independent variables
• b. index composite measure of
  multiple dimensions
• of a particular
  theoretical concept
• gt this is generally a measure
  of your dependent
• variable

4
B. Types of Scales

• 1. Likert Scales
• a. provides independent responses and
  reasonable
• variance within response sets
• b. overall score is based on the
  prescribed
• responses believed to reflect a
  clear picture of a
• particular concept
• c. based on a 4-9 point interval
  measure generally
• ranging from Strongly Disagree
  to Strongly
• Agree
• gt can also respond Never
  through Always

5

• d. procedure
• 1) conceptualize
• 2) specify indicators
  reflecting your concept
• 3) consider the range of human
  emotion
• gt i.e., strong feelings to
  ambivalence
• 4) there must be a rank order
  of response
• a) i.e., how do you feel
  about the question My
• parents (or Research
  Methods professor)
• are always right?
• b) simple Yes or No
  responses
• 5) caution

6

• a) some (Babbie being
  one) argue that it
• matters what you
  label 1 (Strongly agree
• or Strongly Disagree)
  since that may
• influence respondents
  to answer a certain
• way
• b) large number of 3
  responses (Not
• applicable or
  Havent thought about it)
• may invalidate scale
  since there is no
• meaningful strategy
  in which to assess its
• meaning
• c) cure consult the
  literature

7

• 2. Bogardus Social Distance Scale
• a. a composite measure designed to
  examine degree
• or strength of an attitude or
  opinion
• b. What is the right answer?
• c. procedure
• 1) pass through a logical
  sequence of distances
• to bring respondent
  closer to his/her true
• 2) through inference, can be
  a dichotomous
• response set (Yes / No)
  or a Likert-type scale
• response set (Strongly
  agree Strongly
• Disagree)
• d. example

8

• gt How do male police officers actually
  feel about female police
• officers?
• Respond using the following
  answers
• 1) Strongly Agree 2) Agree
  3) Havent thought about it
• 4) Disagree 5) Strongly
  Disagree
• a) I believe that women have a place in
  policing.
• b) I believe that women should be
  hired as police
• officers.
• c) I believe that women should
  have the same duties
• and responsibilities as men.
• d) I believe that a woman officer
  would make a

9
C. Indexes

• 1. indices summarize specific
  observations into a
• representative general usable
  entity
• 2. because dichotomous responses (Yes /
  No) do not
• afford realistic variation or
  information, indexes
• provide a more reasonable / useful
  measure
• 3. e.g. how do you determine whether or
  not someone
• is a criminal?
• a) first must create an operational
  definition
• b) Elliott and Ageton Delinquency
  Index

10
D. Summary

• 1. scales / indexes (complex measures)
  are good
• strategies for determining complex /
  sensitive
• variables, turning aggregate
  information into
• understandable statistics
• gt e.g.s delinquency, crime,
  prejudice, fear of
• victimization, levels
  of corruption...
• 2. multiple indicators allow for more
  thorough statistical
• analysis
• 3. with reasonably conceptualized
  theories, Likert scales
• are good at examining abstract
  concepts
• 4. Bogardus scales are good at measuring
  attitudes
• concerning classes of people or
  things