# Aligning Learners' Expectations and Performance by Learning Analytics Systemwith a Predictive Model

**Authors:** Saša Brdnik, Boštjan Šumak, Vili Podgorelec  
**arXiv:** [2211.07729](https://arxiv.org/abs/2211.07729)  
**Date:** 2022-11-14  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2211.07729v1/

## Key points

- Learning analytics (LA) is a tool for improving student learning and performance by collecting, analyzing, and representing data about learners.
- Adoption of LA in higher education institutions has been slow, with Slovenian institutions lagging behind other European countries.
- Researchers used surveys and focus groups to gather students' expectations of LA in higher education.
- Focus groups allowed for in-depth exploration of students' perceptions and expectations without the limitations of questionnaires.
- Participants were asked about their interest in receiving information through LA, how often they would like access to this information, and whether they would like teaching staff to have access as well.
- Based on these expectations gathered from students, researchers developed a requirement-oriented LA solution consisting of an interface that utilizes grade prediction with an AI model for a selected course.
- The main benefit of this system is its support for self-regulation of the learning process during the semester.
- Positive grade predictions presented students with predicted grades along with visual and text explanations while negative grade predictions prompted learners to change their behavior by comparing their behavior trends with academically successful and unsuccessful students.
- This research highlights the potential benefits of using LA to improve student learning outcomes in higher education institutions.

## Summary

Learning analytics (LA) is a powerful tool for improving student learning and performance by collecting, analyzing, and representing data about learners. The adoption of LA in higher education institutions has been slow, with Slovenian institutions lagging behind other European countries. To address this gap, researchers used a combination of surveys and focus groups to gather students' expectations of LA in higher education. The focus groups allowed for in-depth exploration of students' perceptions and expectations without the limitations of questionnaires. Participants were asked about their interest in receiving information through LA, how often they would like access to this information, and whether they would like teaching staff to have access as well. They were also asked about their feelings regarding data collection, processing, and publication in view of the proposed LA solution. Based on these expectations gathered from students, researchers developed a requirement-oriented LA solution consisting of an interface that utilizes grade prediction with an AI model for a selected course. The interface includes early grade prediction, peer comparison, historical data overview, and self-reported effort in hours spent working on the course. The main benefit of this system is its support for self-regulation of the learning process during the semester. The proposed system's initial evaluation showed positive results among ten students who received access to anonymized data on two representative users via the dashboard. Two user flows were considered in user interface development: one for positive grade predictions and another for negative grade predictions. Positive grade predictions presented students with predicted grades along with visual and text explanations while negative grade predictions prompted learners to change their behavior by comparing their behavior trends with academically successful and unsuccessful students. Overall, this research highlights the potential benefits of using LA to improve student learning outcomes in higher education institutions. By gathering student expectations through qualitative analysis methods such as surveys and focus groups, researchers can develop requirement-oriented solutions that meet student needs while supporting self-regulated learning behaviors throughout the semester. As such systems become more widely adopted, they have the potential to transform higher education by providing personalized learning experiences that help students achieve their academic goals.