Assessment of Test Items with Rasch Measurement Model

Area of Science:

• Educational Measurement
• Psychometrics
• Economics Education

Background:

• Assessing the quality of educational achievement tests is crucial for accurate student evaluation.
• Traditional methods may not fully capture item performance and test reliability.
• The Rasch model offers a robust framework for item analysis and test calibration.

Purpose of the Study:

• To assess the item difficulty index of an Economics Achievement Test using the Rasch model.
• To determine the infit and outfit measures, and overall test reliability.
• To provide recommendations for utilizing the Rasch model in test development.

Main Methods:

• A 100-item Economics Achievement Test was administered to 200 students.
• Data were analyzed using the Winsteps software (version 3.75) implementing the Rasch model.
• Item difficulty, infit/outfit statistics (MNSQ, ZSTD), and reliability (Kuder-Richardson 20) were calculated.

Main Results:

• The difficulty index for each test item was successfully identified.
• Infit and outfit measures were determined, indicating item fit to the Rasch model.
• The overall reliability of the Economics Achievement Test was estimated at 0.86.

Conclusions:

• The Rasch model is effective for assessing item difficulty and ensuring parameter stability in achievement tests.
• The study validates the use of the Rasch model for psychometric analysis of educational instruments.
• Recommendations emphasize the consistent application of the Rasch model for robust test evaluation.