CAT administration of language placement examinations

Area of Science:

• Educational Technology
• Language Assessment
• Psychometrics

Background:

• Traditional language proficiency tests can be lengthy and resource-intensive.
• Computerized testing offers opportunities for innovative assessment designs.
• Adaptive testing frameworks can enhance efficiency and accuracy.

Purpose of the Study:

• To develop and describe a computerized adaptive test (CAT) for language proficiency at Cegep de Jonquiere.
• To investigate the potential of CAT to reduce the number of items required for accurate assessment.
• To explore initial methods for detecting dishonest testing behavior, specifically 'cheating low'.

Main Methods:

• Development of a calibrated item bank for language proficiency assessment.
• Implementation of a computerized adaptive testing algorithm.
• Design and testing of a detection mechanism for 'cheating low' behavior.

Main Results:

• The computerized adaptive test was successfully developed and implemented.
• Adaptive testing allows for accurate proficiency assessment with fewer items compared to traditional tests.
• Initial methods for detecting 'cheating low' were explored, showing potential for identifying dishonest test-takers.

Conclusions:

• Computerized adaptive testing provides an efficient and effective method for language proficiency assessment in an educational setting.
• The developed CAT system has the potential to optimize testing procedures and resource allocation.
• Further research is warranted to refine 'cheating low' detection methods for robust academic integrity measures.