Integrating AI for enhanced student performance analysis and teacher support
Dominguez Eguia, Aaron (2025)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025120934146
https://urn.fi/URN:NBN:fi:amk-2025120934146
Tiivistelmä
This thesis examines the potential integration of an AI tool designed to enhance the analysis of student performance and support educators at The English Montessori School in Madrid. The purpose of the research is to analyze the existing assessment data and reflect on the possibilities and implications of the integration of an AI assessment in Year 6 Maths. More specifically, the study examines how AI-generated assessments and feedback can be utilised to support teachers in analysing student performance, identifying learning gaps, and enhancing the formalisation of formative assessment processes.
The methodology will involve an in-depth examination of student performance data, such as grades, school work and test scores to determine how AI-based analytics can be used to improve educational performance. This paper critically analyzes the available models, literature, and case studies to establish the effectiveness and challenges of AI integration in learning settings.
The findings show that the GPT-5 model had high statistical consistency with teacher ratings, but the feedback was not always contextually comprehended and pedagogically accurate. The AI proved effective and reliable, but over-explained or misunderstood answers, demonstrating that AI is now able to complement, but not substitute, teachers in providing accurate and meaningful assessments.
The integration of AI into education present considerable opportunities for improving the learning process and enabling teachers with data-driven solutions. Future studies must be aimed at improving AI applications in accordance with real-life case studies and increasing their functionality to facilitate personalized education in an effective manner.
The methodology will involve an in-depth examination of student performance data, such as grades, school work and test scores to determine how AI-based analytics can be used to improve educational performance. This paper critically analyzes the available models, literature, and case studies to establish the effectiveness and challenges of AI integration in learning settings.
The findings show that the GPT-5 model had high statistical consistency with teacher ratings, but the feedback was not always contextually comprehended and pedagogically accurate. The AI proved effective and reliable, but over-explained or misunderstood answers, demonstrating that AI is now able to complement, but not substitute, teachers in providing accurate and meaningful assessments.
The integration of AI into education present considerable opportunities for improving the learning process and enabling teachers with data-driven solutions. Future studies must be aimed at improving AI applications in accordance with real-life case studies and increasing their functionality to facilitate personalized education in an effective manner.