This page describes features and settings in VTS that can help increase test security. By high test security, we mean a low risk of test results being influenced by the use of external tools or supporting measures during testing to improve or otherwise influence the test results.
Use the Testplayer Client
-
The Testplayer Client is a desktop application that must be installed locally on the Windows PC on which testing is to be carried out (see Installation of the VTS clients).
-
If it is possible to install the Testplayer Client, testing via this client offers higher test security. The Testplayer Client can help increase test security by restricting certain actions during test administration (e.g. exiting full-screen mode is not possible). It therefore offers greater test security than testing in a browser (as is possible for certain tests).
-
The Testplayer Client cannot be used on mobile devices (smartphones/tablets). It therefore cannot be used for testing conducted via invitation links.
-
In the VTS settings, you can enforce the use of the Testplayer Client for testing that you start directly from the VTS. To do this, go to Settings → Testing → General → Test administration and security. There you will find the option Test can be carried out in a browser. Deactivate this option, if you wish to enforce the use of the Testplayer Client
Please note: If you use invitation links (for open mode or proctoring), the Testplayer Client will not be used (even if this option is not enabled). However, use of the Testplayer Client is possible for proctoring.
-
The Testplayer Client can enhance control over the test setting and reduce certain browser-based risks. However, it does not, on its own, prevent the use of external devices, such as a smartphone or a second computer or screen. To ensure a high level of test security, it should therefore be combined with organisational measures (such as supervising the testing).
Taking the test under supervision
-
Where possible, administer the test under the supervision of a trained test administrator (controlled mode or proctored mode) and ensure the necessary conditions are in place (e.g. no mobile phones, smart devices, etc. permitted during the testing).
Use open mode testing only for screening
-
If you use ability tests in open mode, the results should only be used for screening purposes and must be verified by a second supervised test (see DIN 33430, Chapter 6.2; Deutsches Institut für Normung e. V., 2016). This applies in particular to vocational aptitude testing, but also to other decisions with relevant consequences for the test taker.
Inform the test takers
-
Inform the test takers about the purpose, procedure, conditions and requirements of the testing. This reduces the perceived pressure to resort to external aids for test preparation.
-
Make it explicitly clear to test takers that the use of AI applications, chatbots, translation functions or other external aids is not permitted during testing, unless these have been expressly authorized.
Check the test protocols in the VTS
-
In the test results, you can use the test protocols or item analysis protocols to track which answer the person gave to each test item and how long it took them to answer.
-
Unusual response patterns (e.g. very consistent or highly fluctuating working times, stereotypical response patterns) could indicate irregularities during testing. Such anomalies should not be interpreted in isolation, but may warrant a plausibility check, further enquiry or a retry of testing under controlled conditions.
Use time-limited tests
-
Using time-limited tests makes it more difficult for test takers to use AI.
Use non-verbal item material
-
Consider using tests (or subtests of tests) that include non-verbal test materials (e. g. figural matrices and other visual-spatial or interactive items), especially for open mode testing. Such materials are harder to copy into external tools (like AI chatbots) and less likely to be successfully solved by AI than purely verbal test materials (see Abdelkarim et al., 2025).
References
Abdelkarim, S., Lu, D., Flores, D.-L., Jaeggi, S., & Baldi, P. (2025). Evaluating the intelligence of large language models: A comparative study using verbal and visual IQ tests. Computers in Human Behavior: Artificial Humans, 5, Article 100170. https://doi.org/10.1016/j.chbah.2025.100170
Deutsches Institut für Normung e. V. (2016). DIN 33430:2016-07: Anforderungen an berufsbezogene Eignungsdiagnostik. DIN Media.