Rejoinder to McNeish and Mislevy: What Does Psychological Measurement Require?

Area of Science:

• Psychometrics
• Psychology

Background:

• Responded to critiques on the merits of the simple sum score.
• Highlighted the need for psychometricians to educate researchers on psychometric models.
• Addressed issues regarding test construction methods and data quality.

Purpose of the Study:

• To advocate for the responsible use of latent variable scores and sum scores.
• To emphasize the role of psychometrics education in bridging the gap with psychology.
• To discuss the practical communication of test results to practitioners and clients.

Main Methods:

• Rejoinder to previous commentaries.
• Discussion of Item Response Theory (IRT) models.
• Analysis of test and questionnaire construction methods.

Main Results:

• Different test construction methods may yield similar results unless attribute theory is unarticulated, leading to noisy data.
• Sum scores and their transformations (e.g., percentiles) facilitate communication of test results.
• Latent variables are crucial for advanced psychometric applications like equating and adaptive testing.

Conclusions:

• A significant gap exists between psychology and psychometrics that requires collaborative efforts to bridge.
• Psychology and psychometrics must work together to improve the application and interpretation of test scores.
• While advanced applications use latent variables as technical tools, simpler scores aid communication in practical decision-making.