Computer Assisted Testing Basis Explained

Computer Assisted Testing (CAT) refers to assessment systems where a computer administers tests and adapts the difficulty of questions based on the test-taker's performance.

Role of Item Response Theory (IRT)

The core principle enabling CAT's adaptive nature is Item Response Theory (IRT). IRT models provide a mathematical framework to:

• Estimate a test-taker's ability level.
• Estimate the difficulty and other characteristics of each test item.
• Select the next item to be administered based on the current ability estimate, aiming for maximum information and precision.

This allows the computer to efficiently tailor the test to the individual, providing a more accurate measure of ability compared to fixed tests.

Evaluating Other Options

• Efficiency of modern computers: While essential for CAT operation, computer efficiency is a technological enabler, not the underlying psychometric principle.
• Emphasis on internal consistency: Internal consistency measures (like Cronbach's alpha) assess reliability but do not directly guide item selection or adaptation in CAT.
• Compatibility between manual and computerised scoring: This relates to scoring methods, not the adaptive testing logic itself.

Therefore, Item Response Theory is the foundational concept for Computer Assisted Testing.