ABSTRACT
Social and behavioral scientists often analyze ordinal outcomes, which are variables with categories ranked from low to high on a single dimension or characteristic. However, the distances between categories are unknown. For example, the General Social Survey (GSS) includes a measure of attitudes toward same-sex marriage based on the following question: “Do you agree or disagree … homosexual couples should have the right to marry one another.” The response options are based on a Likert scale: 1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, and 5 = strongly agree. The categories are ranked from low to high according to the level of support for same-sex marriage, but it is unclear whether the distances between adjacent categories are equal. Survey researchers typically assign the values 1-5 for a Likert scale question with five categories, but the difference between categories 1 and 2 is not necessarily the same as the difference between categories 2 and 3 and so forth. Additionally, ordinal variables may have multiple dimensions. In this example, the categories reflect agreement/disagreement but also the intensity of opinion: strongly versus somewhat.
