Controlling the Minimum Item Exposure Rate in Computerized Adaptive Testing: A Two-Stage Sympson-Hetter Procedure.

Area of Science:

• Psychometrics
• Educational Measurement
• Computerized Adaptive Testing (CAT)

Background:

• Computerized adaptive testing (CAT) enhances efficiency but leads to unbalanced item usage.
• Uneven item exposure rates create test security risks (overexposure) and economic issues (underexposure).

Purpose of the Study:

• To introduce and evaluate a novel two-stage Sympson-Hetter (TSH) method for balanced item pool utilization in CAT.
• To simultaneously control minimum and maximum item exposure rates to address over/underexposure issues.

Main Methods:

• The TSH method divides CAT into two stages: Stage 1 increases exposure of underexposed items above a minimum rate (r_min).
• Stage 2 controls exposure of all items below a maximum rate (r_max) to prevent overexposure.
• TSH uses a minimal number of less-discriminating items in Stage 1 to mitigate impact on trait estimation.

Main Results:

• The TSH method significantly improves item pool usage.
• Trait estimation precision is largely maintained under most simulation conditions.
• Required levels of test security are preserved.

Conclusions:

• The TSH method offers an effective solution for balanced item pool utilization in CAT.
• It addresses critical issues of item overexposure and underexposure.
• TSH provides a viable approach for enhancing CAT psychometric and operational properties.