Computer Adaptive Testing
The Q-LEVEL language tests are using computer adaptive test technology. In contrast to conventional language tests, in which all candidates complete the same fixed set of tasks, adaptive tests use a flexible set of tasks for each test candidate, in terms of the number and type of task.
The test system adapts to the test person’s level of ability. After a successfully completed task, the algorhythm assigns a more difficult task to the test person and an easier one after an incorrectly completed task. The test breaks off at the level of difficulty at which the test candidate just completed the task correctly and does not manage to complete the next more difficult task.
Exact and Fast Thanks to Adaptive Algorhythm
Each of the four sub-tests vocabulary, grammar, reading and listening builds on the previous one.
The starting point for taking a Q-LEVEL test is vocabulary exercise at level B1, the third of six CEFR levels. The language level that is achieved in the Vocabulary subtest is the starting point for the Grammar subtest and the mean of the Vocabulary and Grammar results in turn defines the level at which the Reading subtest begins. In this way, both the global competence as well as the competences for vocabulary, grammar, reading and listening can be tested in a very short time.
For each online language test, a corpus (item bank) of >300 calibrated items, scale to the CEFR levels with a total of >1200 answer options is available.
Read more about the scientific base for computer adaptive testing here (in German).
- Determination of the target level in a very short time
- Broadly impossible to set an over-challenging or under-challenging task
- High test safety in group tests (cheating is impossible due to different tasks)
- Multiple test participations possible (placement, large-scale learning progress, etc.)