Technical information :: statistical evaluation of scales
Evaluation of scales is a process whereby evidence is accumulated and summarised in the quality framework
What is the meaning of validity?
The most essential characteristic of a scale is that it measures what it claims to measure. A physical attribute, height for example, can be measured very accurately and everybody trusts that a ruler measures what it says. In contrast, a psychological state, dependence for example, is much more difficult. The quality framework rates several different types of validity.
What is the meaning of reliability?
A measure can be reliable without being valid but a valid measure must always be reliable, meaning that people respond to the questionnaire items in the same way in different circumstances.
Finding the structure of a scale?
It is important to understand what a scale is purported to measure. The Leeds Dependence Questionnaire, for example, measures the single construct of dependence. The Addiction Recovery Questionnaire, for example, measures recovery but has three subscales. During scale development, factor analysis is commonly used to determine which items cluster together and measure a particular factor, which may then be defined as a subscale.
How do you test the validity of scales?
The quality framework evaluates validity using a number of different statistical techniques all of which depend on some form of correlation.