Scaling, Guttman

Example

A well-known example using a Guttman scale is the Bogardus social distance scale. It assesses the extent to which people are willing to interact with members of other groups. For example, imagine one were interested in measuring people’s willingness to interact with Muslims. The scale could include statements such as:

• 1.
  I am comfortable with Muslims living in the United States. (least willingness to interact)
• 2.
  I am okay with Muslims living in my city.
• 3.
  I am comfortable with Muslim families in my neighborhood.
• 4.
  I am okay with having Muslims as acquaintances.
• 5.
  I am comfortable working beside Muslims in the workplace.
• 6.
  I am fine with having Muslims as friends.
• 7.
  I am comfortable having Muslims as close family members. (most willingness to interact)

The statements reflect progressively more willingness to interact with Muslims from the first statement to the seventh statement. If someone agrees with the fourth statement, he or she has implicitly agreed with statements 1, 2, and 3, the rank-ordering of the statement. Similarly, agreement with statement 7 implies one agrees with all previous statements. Knowing the last statement a respondent agreed with offers information about what other statements he or she has agreed with, or his or her pattern of agreement.

A cumulative score (a single number) can be computed that reflects the person’s overall willingness to interact with Muslims. It is cumulative in the sense that agreement cumulates, or builds up, from the first sentence to the last one the respondent agreed with. For example, the cumulative score could be a 7 for a person who has agreed with all seven statements, a 4 for a person who has agreed with the first four statements, and so on. One of the main goals of Guttman scaling is to enable a researcher to predict responses to individual items or statements based on a respondent’s cumulative score. In other words, knowing someone’s cumulative score for the statements is a 4 means one can predict the person would agree with statements 1, 2, 3, and 4, but not to statements 5, 6, and 7.

Development

There are several steps that one would undertake when developing a Guttman scale to measure a selected concept of interest. Imagine one were interested in assessing people’s willingness to allow vaping in public places. The first step is to develop a large number of statements that could capture whether people have favorable attitudes or not toward vaping in public. Researchers sometimes develop such statements by themselves, by consulting others in their discipline, or by consulting a panel of experts on the topic of interest. Examples of statements could include, “I don’t have a problem with people vaping in public” or “I am absolutely against people being allowed to vape in public.”

...