The AI marks the way an experienced human rater would – looking at several subskills and combining them into an overall score:
The AI has been trained on millions of data points over eight years, with continual human calibration and user testing.
In 2024, the AI and a panel of experienced human raters independently scored the same set of student responses. Human and automated scores showed a very high level of consistency:
The scores given by the group of experts aligned closely with those generated by the AI system. There was a high degree of agreement between human evaluators and the AI technology.
The Dynamic Speaking Test has been shaped by four prime movers: ClarityEnglish, the publisher, and three independent specialists in language testing and assessment.