Pre-equating: a simulation study based on a large scale assessment model.

Area of Science:

• Educational Measurement
• Psychometrics
• Item Response Theory

Background:

• Post-equating (PE) is standard for score equating, but time constraints can prevent its use.
• Pre-equating (PrE), utilizing existing item bank calibrations, is a potential alternative.
• Assessing PrE feasibility requires comparing calibration equivalency and error against PE.

Purpose of the Study:

• To evaluate the feasibility of pre-equating (PrE) as an alternative to post-equating (PE).
• To examine the impact of item calibration stability on PrE and PE outcomes.
• To analyze measurement error and examinee placement differences between PrE and PE.

Main Methods:

• Item banks were created over three periods of item introduction.
• The Rasch model was employed to analyze item parameter recovery and measurement error.
• Data were analyzed for both pre-equating (PrE) and post-equating (PE) scenarios.

Main Results:

• Pre-equating (PrE) is a viable alternative to post-equating (PE) under specific conditions.
• Large sample sizes are crucial for enhancing item calibration stability in the item bank.
• Item calibration stability directly impacts the accuracy of examinee placement in PrE.

Conclusions:

• Pre-equating (PrE) can be successfully implemented when time constraints preclude post-equating (PE).
• The stability of item calibrations, bolstered by large sample sizes, is paramount for PrE efficacy.
• This study supports PrE as a practical solution in educational and psychological testing.