A Multi-Source Dataset for CS1 Failure Prediction

	Description Demographics, LMS logs, surveys and continuous assessments from four cohorts At a glance 284 students across four cohorts · 87,543 Moodle log events · Surveys and continuous assessment records Author Lighton Phiri Abstract This dataset contains 284 students enrolled in a CS1 course (Computer Systems and Architecture) across four academic cohorts (2017/18–2020/21) at the University of Zambia. It integrates four data sources: student information system records, a pre-course survey, Moodle LMS logs, and continuous assessment records. This is the first open dataset for CS1 failure prediction from a sub-Saharan African university and is ideal for research in learning analytics, educational data mining, early warning systems, and academic momentum modelling. Year of Publication 2026 Repository Kaggle Data https://www.kaggle.com/datasets/lightonphiri/a-multisource-dataset-for-cs1-failure-prediction DOI 10.34740/KAGGLE/DSV/17913574 Licence CC BY 4.0 Google Scholar DOI BibTeX

Description

Demographics, LMS logs, surveys and continuous assessments from four cohorts

At a glance

284 students across four cohorts · 87,543 Moodle log events · Surveys and continuous assessment records

Author

Abstract

This dataset contains 284 students enrolled in a CS1 course (Computer Systems and Architecture) across four academic cohorts (2017/18–2020/21) at the University of Zambia. It integrates four data sources: student information system records, a pre-course survey, Moodle LMS logs, and continuous assessment records. This is the first open dataset for CS1 failure prediction from a sub-Saharan African university and is ideal for research in learning analytics, educational data mining, early warning systems, and academic momentum modelling.

Year of Publication

2026

Repository

Kaggle

Data

https://www.kaggle.com/datasets/lightonphiri/a-multisource-dataset-for-cs1-failure-prediction

DOI

10.34740/KAGGLE/DSV/17913574

Licence

CC BY 4.0