Criterion validity, construct validity, and factor analysis: An introductory overview

Area of Science:

• Psychometrics
• Health Sciences Research Methodology

Background:

• This article is the final in a series on developing and psychometrically testing rating scales.
• It builds upon previous discussions of content validity, focusing on criterion and construct validity.

Purpose of the Study:

• To detail the assessment of criterion validity and construct validity in rating scales.
• To explain the application of factor analysis (FA) as a method for evaluating construct validity.

Main Methods:

• Criterion validity is assessed by correlating a new scale with a "gold standard" criterion (concurrent or predictive).
• Construct validity is evaluated by examining convergent and divergent validity, often using Pearson's correlation coefficient.
• Exploratory factor analysis (FA) steps are detailed as a multivariate technique to assess relationships between variables and underlying factors.

Main Results:

• Pearson's correlation coefficient and phi coefficient are presented as measures for criterion validity.
• Pearson's correlation coefficient is used to estimate convergent and divergent validity for construct validity.
• Factor analysis is identified as a key method for assessing construct validity, with exploratory FA discussed in detail.

Conclusions:

• Accurate psychometric testing, including criterion and construct validity, is crucial for reliable rating scales.
• Factor analysis provides a robust method for evaluating the underlying structure and construct validity of measurement tools.
• Understanding these validation techniques ensures the development of high-quality instruments for research and clinical practice.