Economic Strain and Depressive Symptoms Among 500 University Students in Bangladesh

An anonymized analytic dataset of 500 undergraduate students from Bangladesh, used to study the relationship between economic strain and depressive symptoms. The dataset is accompanied by a full Jupyter Notebook replication script for data preprocessing, descriptive statistics, Ordinal Logistic Regression, Structural Equation Modeling, and XGBoost classifier visualizations with SHAP values. The data was deposited by Muhammad Hassan Bin Afzal of Jahangirnagar University and is hosted on Mendeley Data.

Use Cases

• Replicate Ordinal Logistic Regression models based on economic strain and depressive symptom variables mentioned in the description.
• Perform Structural Equation Modeling to analyze fit indices for quality-of-life pathways.
• Train and explain multiclass XGBoost classifiers using permutation-based SHAP values for feature importance.
• Conduct data preprocessing and generate descriptive statistics for a survey-based student population.

Strengths

• Includes a full Jupyter Notebook replication script that reproduces all analysis from the published study.
• Dataset contains records for 500 undergraduate students, providing a specific sample size.
• Analysis includes multiple modeling techniques: Ordinal Logistic Regression, Structural Equation Modeling, and XGBoost.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the analytic dataset is known (500), but other metadata like file formats and license are unknown.
• Data reflects a specific geographic and demographic bias, focusing solely on university students in Bangladesh.

Provenance

Source

Muhammad Hassan Bin Afzal, Department of Statistics and Data Science, Jahangirnagar University.

Collection Method

Likely collected via survey from undergraduate students.

Time Range

null

Freshness

Last updated 2026-05-04 23:04:33; freshness should be verified.

Geography

Bangladesh