ag00dman/student-depression-analysis

---
license: mit
task_categories:
- tabular-classification
tags:
- mental-health
- education
- economics
---

# **Assignment #1: EDA & Dataset**
## **Predicting and Preventing Student Depression**
**Student:** Amit Goodman
**Program:** Economics & Entrepreneurship, Reichman University (RUNI)
**Date:** March 2026

### **Project Overview**
In this project, I explore the "Student Depression Dataset" to build a narrative around student well-being. By analyzing academic pressure, financial stress, and lifestyle habits, I aim to identify predictable risk factors and uncover actionable protective measures.

### **Data Handling & Quality Protocol**
To ensure a high-quality analysis, I developed a strict cleaning protocol executed directly within the project notebook. While the source CSV remains in its original state, the following steps are performed via code to prepare the data for analysis:

* **Outlier Removal:** I identify and remove rows with a 0.0 CGPA, treating them as data entry errors.
* **Domain Filtering:** I restrict the dataset to the **Student** profession to keep the focus contextually relevant to academic life.
* **Missing Value Management:** I address the negligible 0.01% of missing values in `Financial Stress` and drop irrelevant identifiers like `id`.

### **Exploratory Data Analysis (EDA)**

#### **1. Defining the Problem**
I began by analyzing the distribution of depression within the student population to understand the scope of the challenge.


#### **2. The Primary Risk Factors: Pressure & Finance**
I focused on the intersection of academic and economic stress. My analysis shows a direct correlation between these stressors and mental health outcomes.


**Key Insight:** `Academic Pressure` and `Financial Stress` emerged as the most significant predictors, showing a high density of depression cases at the upper levels of these scales.

#### **3. The Protective Buffers: Satisfaction & Lifestyle**
I looked for "Actionable Measures" - factors that can mitigate these risks.


### **The Resolution (Conclusion)**
The data reveals that the relationship between lifestyle and mental health is more complex than a simple "more is better" rule.

**My Key Findings:**
* **The High-Risk Extreme:** Students sleeping **less than 5 hours** are at the absolute highest risk of depression.
* **The Sleep Paradox:** Interestingly, the data indicates that students in the **7-8 hour** bracket actually have a higher rate of depression than those in the **5-6 hour** bracket. This suggests that in this specific population, standard sleep duration might correlate with other depressive symptoms like lethargy.
* **Satisfaction Matters:** High **Study Satisfaction** acts as a primary protective factor, even when academic pressure is high.
* **Actionable Advice:** To lower the statistical risk of depression, my findings suggest prioritizing a healthy diet, ensuring high alignment with one's academic path, and avoiding the "extreme low" of sleep (<5 hours).

### **Files in this Repository**
* `Student_Depression_Dataset.csv`: The raw, original dataset used for this project.
* `Assignment_1_EDA_&_Dataset_Amit_Goodman.ipynb`: The complete Python analysis, including the **data cleaning scripts**, visualizations, and insights.

license: MIT许可证
task_categories:
- 表格分类（tabular-classification）
tags:
- 心理健康
- 教育
- 经济学

# **作业#1：探索性数据分析（Exploratory Data Analysis，EDA）与数据集**
## **预测与预防学生抑郁**
**学生：阿米特·古德曼（Amit Goodman）**
**专业：经济学与创业学，赖克曼大学（Reichman University，RUNI）**
**日期：2026年3月**

### **项目概述**
本项目围绕“学生抑郁数据集（Student Depression Dataset）”展开，旨在构建学生福祉相关的研究叙事。通过分析学业压力、经济压力与生活习惯，本研究旨在识别可预测的风险因素，并挖掘可落地的防护措施。

### **数据处理与质量管控流程**
为保障分析质量，本项目制定了严格的数据清洗流程，并直接在项目笔记中执行。尽管源CSV文件保持原始状态，但我们通过代码完成以下数据预处理步骤，以适配分析需求：
* **异常值剔除**：识别并移除平均学分绩点（Cumulative Grade Point Average，CGPA）为0.0的样本行，将其视为数据录入错误。
* **领域筛选**：将数据集限定为**学生群体**职业类别，确保分析聚焦于学术生活场景。
* **缺失值处理**：针对`Financial Stress`（经济压力）字段中占比仅0.01%的少量缺失值进行处理，并移除`id`等无关标识符。

### **探索性数据分析（Exploratory Data Analysis，EDA）**
#### **1. 问题界定**
首先分析学生群体中的抑郁分布情况，以明确该问题的影响范围。

#### **2. 核心风险因素：压力与经济状况**
本研究聚焦于学业压力与经济压力的交互影响，分析结果显示这两类压力与心理健康结局存在直接关联。
**核心发现**：`Academic Pressure`（学业压力）与`Financial Stress`（经济压力）成为最具影响力的预测因子，在这两类压力量表的高分区间，抑郁案例的分布密度显著更高。

#### **3. 防护缓冲因素：满意度与生活方式**
本研究探寻了“可落地干预措施”，即能够缓解上述风险的影响因子。

### **研究结论**
数据分析显示，生活方式与心理健康之间的关联并非简单的“越多越好”，而是更为复杂。
**核心研究结论**：
* **极端高风险群体**：日均睡眠**不足5小时**的学生抑郁风险最高。
* **睡眠悖论**：值得注意的是，数据显示睡眠时长处于**7-8小时**区间的学生，抑郁发生率反而高于**5-6小时**区间的学生。这表明在该研究群体中，标准睡眠时长可能与嗜睡等其他抑郁症状存在关联。
* **满意度的作用**：较高的**学业满意度**是核心防护因素，即便在学业压力较高的情况下亦是如此。
* **可落地建议**：为降低抑郁的统计风险，本研究建议优先保障健康饮食、确保与自身学术路径高度适配，并避免睡眠时长处于“极端低值”（<5小时）。

### **本仓库包含文件**
* `Student_Depression_Dataset.csv`：本项目使用的原始未处理数据集。
* `Assignment_1_EDA_&_Dataset_Amit_Goodman.ipynb`：完整的Python分析代码，包含**数据清洗脚本**、可视化内容与研究结论。