Abstract
The author examined critically three beliefs of S.S. Stevens pertaining to his quadripartite system of scales of measurement:
(1) There are four scales of measurement in common usage (2) These scales and the scientific disciplines that use them can
be incrementally graded for levels of reliability and validity or accuracy such that: Nominal scale variables produce the
lowest levels of reliability and accuracy; with successively improving levels for Ordinal, Equal Interval, and Equal Ratio
scales; and (3) The scale upon which a variable is measured determines the type of statistical test that one is permitted
to apply. It was shown that each of the three beliefs is fundamentally flawed.
(1) There are four scales of measurement in common usage (2) These scales and the scientific disciplines that use them can
be incrementally graded for levels of reliability and validity or accuracy such that: Nominal scale variables produce the
lowest levels of reliability and accuracy; with successively improving levels for Ordinal, Equal Interval, and Equal Ratio
scales; and (3) The scale upon which a variable is measured determines the type of statistical test that one is permitted
to apply. It was shown that each of the three beliefs is fundamentally flawed.
- Content Type Journal Article
- Category Original Paper
- Pages 1-7
- DOI 10.1007/s10803-012-1486-z
- Authors
- Domenic V. Cicchetti, Child Study Center, Yale University School of Medicine, New Haven, CT 06511, USA
- Journal Journal of Autism and Developmental Disorders
- Online ISSN 1573-3432
- Print ISSN 0162-3257