ItemAnalyse

Itemanalyse is the systematic examination of individual test items to evaluate their quality and function within a measurement instrument. In educational and psychological testing, it helps determine whether items behave as intended and contribute to reliable, valid measurement. The goal is to improve the overall quality of a test by identifying and revising or removing problematic items.

In classical test theory (CTT), item analysis commonly reports item difficulty, measured as the proportion of respondents answering correctly (the p-value); item discrimination, often assessed by the point-biserial correlation between item score and total test score; and item-total correlations. Distractor analysis considers how often each incorrect option is chosen, guiding revisions of multiple-choice items. Items with very low or very high difficulty or weak discrimination are candidates for revision or removal.

In item response theory (IRT), analysis centers on item parameters such as difficulty, discrimination, and guessing. Item characteristic curves and information functions describe how item performance varies with the latent ability being measured, enabling more precise measurement across different levels of ability.

Applications and limitations: itemanalyse supports test refinement, shorter tests with preserved accuracy, and fairness monitoring. It requires adequate sample sizes and representative populations, and results depend on the characteristics of the group tested. Fairness checks such as differential item functioning (DIF) analysis help identify biased items. The process is iterative: items are revised and re-tested before final use to ensure robust measurement properties.