Adaptive writing feedback
Collaboration:
Prof. Dr. Claudia Schrader (Lehren und Lernen mit digitalen Medien)
Prof. Dr. Stefanie Frisch (Goethe-Universität Frankfurt)
Dr. Stephan Tilgner (KI4BUW - DigiData)
The aim of this particular collaboration project "Adaptive writing feedback" is on the development and implementation of automatically generated feedback on texts written by pupils in the foreign language English on tablets. Parameters of formal and content-related text quality (e.g., spelling, lexis, grammar, coherence, task fulfilment and ingenuity) are continuously tracked and analysed using AI methods. In addition, pressure parameters of the used digital pen during writing are assessed and used as indicators of learning-centred emotions. The relationship between both - cognitive and affective – factors will be examined in order to develop and use automatically generated feedback that is adaptive to the learning performance and the emotional state of each of the individual pupils. Based on the experimental evaluation of various feedback design combinations, recommendations for feedback design will be formulated in order to support pupils’ self-concept, to balance suboptimal emotional states and, as a result, to optimally support the learning process and outcome.
