Using Confidence Intervals of the Item and Test Information Functions to Test Differential Item and Test Functioning:...

Area of Science:

• Educational Measurement
• Psychometrics
• Statistics

Background:

• Differential item and test functioning are crucial for fair assessments.
• Traditional methods for evaluating differential functioning often rely solely on item difficulties.

Purpose of the Study:

• To visualize group differences using confidence intervals of item and test information functions.
• To statistically compare these functions for evaluating differential item and test functioning.

Main Methods:

• Utilized 95% confidence intervals of item and test information functions.
• Estimated item and test information functions, conditional standard errors, and reliability for male and female groups.
• Employed population-based Z-tests and Monte Carlo simulations for statistical confirmation.

Main Results:

• Visualizing confidence intervals of information functions clearly showed group differences.
• Statistical tests confirmed these visual findings, validating the approach.
• The proposed method demonstrated increased sensitivity compared to traditional techniques.

Conclusions:

• Information functions combined with inferential tests provide a robust method for detecting differential group behavior.
• Bootstrap distributions enable the construction of inferential tests for significance.
• This procedure, incorporating item difficulties and discrimination indices, enhances the evaluation of assessment fairness.