Metric invariance between girls and boys in educational assessments: how is it analyzed? How can its evaluation be optimized?
Researchers from the Psychometrics Lab at the University of Granada (@PsychometricUGR), part of the Human Behavior Modeling and Measurement Research Group (HUM-624), have published a mixed systematic review of the scientific literature to investigate the most suitable methodology for analyzing ‘metric invariance’ in educational assessments between girls and boys.
It is essential to establish metric invariance to ensure that an assessment instrument, such as a test or questionnaire, is measuring the same ‘construct’ in different groups, ensuring that the comparisons made of the measures are valid and fair, and not due to ‘artifacts’ in the measurement process. In the case of educational measures, making fair score interpretations between girls and boys is important to ensure equal access to education and educational policies.
The mixed systematic review compiles different strategies for analyzing invariance, both quantitative: structural equations, psychometric models, etc., and qualitative: focus groups, interviews, etc., identifying their advantages and disadvantages, and the evaluation contexts in which they are most frequently used. Additionally, it discusses which invariance analysis strategies are most suitable for investigating the causes of lack of invariance, a crucial issue when designing strategies to ensure the validity and fairness of assessments.
Full reference:
Navarro-González, M.C., Padilla, J.L., & Benítez, I. (2024). Analyzing Measurement Invariance for Studying the Gender Gap in Educational Testing: A Mixed Studies Systematic Review. European Journal of Psychological Assessment. DOI: 10.1027/1015-5759/a000820
Contact researchers:
M. Carmen Navarro-González (mcnavarro@ugr.es)
José-Luis Padilla (jpadilla@ugr.es)
Isabel Benítez (ibenitez@ugr.es)