“hypothetical constructs such as verbal, quantitative, and analytical ability are more usefully interpreted if they can be shown empirically to be measuring different things” 

University of Michigan - Ann Arbor

- Psychometric bias
- Convergent validity
- Discriminant validity
- Construct validation
- Invariant construct validity

Construct Validity of the GRE Aptitude Test Across Populations - An Empirical Confirmatory Study: Definitions of Key Terms

Rock, D. A., Werts, C., & Grandy, J. (1981). CONSTRUCT VALIDITY OF THE GRE APTITUDE TEST ACROSS POPULATIONS‐AN EMPIRICAL CONFIRMATORY STUDY. ETS Research Report Series, 1981(2), i-44. https://onlinelibrary.wiley.com/doi/abs/10.1002/j.2333-8504.1981.tb01284.x 

Construct Validity of the GRE Aptitude Test Across Populations - An Empirical Confirmatory Study

- “does not require external criteria information for the analysis”
- “a first step in a broader process of an integrated validation procedure that should include not only internal checks on the population invariance of the underlying constructs but also checks on the population invariance of their relationship with external criteria.” 

Construct Validity of the GRE Aptitude Test Across Populations - An Empirical Confirmatory Study: Psychometric Bias

“the item types that are assumed to be measures of a hypothetical construct such as analytical ability should demonstrate proportionately higher interrelationships among themselves than with measures of other constructs such as verbal or quantitative ability”

Construct Validity of the GRE Aptitude Test Across Populations - An Empirical Confirmatory Study: Convergent Validity

- “the basic prerequisite to proper interpretation of a test score”
- “the process of marshalling evidence of relationships with other variables to support the inference that an observed test score has a particular meaning”
- “for example, that it is a valid measure of developed verbal or mathematical ability”

Construct Validity of the GRE Aptitude Test Across Populations - An Empirical Confirmatory Study: Construct Validation

- “similar patterns of loadings across populations”
- “equal units of measurement across populations”
- “equal test score precision as defined by the standard error of measurement” 
- Need to consider potential psychometric bias if the 3 criteria are different across the 4 populations 

Learn Before

Related