University Student Stress Dataset

This dataset contains responses from 3000 undergraduate students studying at public, private, and national universities in Bangladesh. The primary goal of the dataset is to support machine learning research on classifying student stress levels using academic, lifestyle, social, and psychological factors. Data were collected through a structured Google Form survey, which was distributed via email to 3710 students, and a total of 3000 complete and anonymous responses were used in the final dataset.

The dataset includes 18 attributes, selected based on their relevance to academic stress in higher education. Demographic features include Age (19–24) and Gender, while University_Type captures differences between institutional environments. Academic indicators—Study_Hours, Class_Attendance, Exam_Frequency, Assignment_Load, and Tuition (Yes/No)—reflect students’ academic workload and overall study pressure.

Lifestyle-related features include Sleep_Hours, Social_Media_Use, Screen_Time, and Physical_Exercise, as daily habits often influence stress and time management. Socio-economic and psychological attributes—Family_Income_Level, Peer_Pressure, Family_Support, and Anxiety_Level—provide additional context on student wellbeing and external influences.

A derived variable, Stress_Score, was computed by combining workload indicators, psychological ratings, and lifestyle habits. Based on this score, each student was assigned a target label, Stress_Level, categorized as Low, Medium, or High. This label enables supervised learning tasks and stress prediction modeling.

The dataset is clean, structured, and suitable for multi-class classification, correlation analysis, and educational data science research. It offers a valuable resource for understanding how academic demands and personal habits interact to influence stress among university students in Bangladesh.