Improving Students Argumentation Learning with Adaptive Self-Evaluation Nudging

Recent advantages from computational linguists can be leveraged to nudge students with adaptive self evaluation based on their argumentation skill level. To investigate how individual argumentation self evaluation will help students write more convincing texts, we designed an intelligent argumentation writing support system called ArgumentFeedback based on nudging theory and evaluated it in a series of three qualitative and quaxntitative studies with a total of 83 students. We found that students who received a self-evaluation nudge wrote more convincing texts with a better quality of formal and perceived argumentation compared to the control group. The measured self-efficacy and the technology acceptance provide promising results for embedding adaptive argumentation writing support tools in combination with digital nudging in traditional learning settings to foster self-regulated learning. Our results indicate that the design of nudging-based learning applications for self-regulated learning combined with computational methods for argumentation self-evaluation has a beneficial use to foster better writing skills of students.