Unterscheidungsleistung

Unterscheidungsleistung, often translated as "discriminative power" or "distinguishing ability," is a concept primarily used in psychometrics and educational testing. It refers to the degree to which an item, such as a test question or a psychological scale item, can differentiate between individuals who possess a high level of the trait or construct being measured and those who possess a low level. In essence, it's a measure of how well an item separates the high scorers from the low scorers on the overall test.

A high Unterscheidungsleistung indicates that the item is effective in identifying individuals who truly understand the material or exhibit the trait. Conversely, an item with low Unterscheidungsleistung may be poorly constructed, too easy, too difficult, or ambiguous, failing to discriminate meaningfully between different ability levels.

This metric is crucial for item analysis, a process used to evaluate and improve the quality of test items. By calculating the Unterscheidungsleistung for each item, test developers can identify and remove or revise problematic items, leading to a more reliable and valid assessment tool. Various statistical methods, such as item discrimination indices (e.g., point-biserial correlation), are employed to quantify Unterscheidungsleistung. A higher positive value generally signifies better discriminative power.