授業の目的 【日本語】
Goals of the Course(JPN) | | | 本ゼミの目的は、Pythonを用いたデータ分析を通じて、社会やビジネスにおける「因果関係」を適切に検証するための基礎的な考え方と実践的スキルを身につけることである。受講生は、相関と因果の違いを理解し、効果検証のためのデータの扱い方、分析手法、および結果の解釈方法を修得する。 |
|
|
授業の目的 【英語】
Goals of the Course | | | The goal of this seminar is to develop a fundamental understanding and practical skills for evaluating causal effects using Python-based data analysis. Students will learn to distinguish correlation from causation and acquire the ability to handle data, apply appropriate analytical methods, and interpret results in the context of impact evaluation in social and business settings. |
|
|
到達目標 【日本語】
Objectives of the Course(JPN) | | 本ゼミ終了時に、学生は以下を達成できるようになる。
1. 効果検証における因果関係とバイアスの基本概念を説明できる
2. A/BテストおよびDIDの基本的な考え方と仮定を理解できる
3. Pythonを用いて簡単な効果検証分析を実装できる
4. 分析結果を解釈し、手法の限界を議論できる
By the end of the seminar, students will be able to:
1. Explain basic concepts of causality and bias in impact evaluation.
2. Understand the logic and assumptions of A/B testing and DID.
3. Implement simple impact evaluation analyses using Python.
4. Interpret empirical results and discuss methodological limitations. |
|
|
授業の内容や構成
Course Content / Plan | | Week 1 – Course Introduction
Overview of impact evaluation and Python setup
Week 2 – What is Impact Evaluation?
Causality, counterfactuals, and bias (Ch.1)
Week 3 – Logic of A/B Testing
Treatments, effects, and experimental design (Ch.2.1–2.5)
Week 4 – A/B Test Analysis in Python
Data analysis and implementation (Ch.2.6)
Week 5 – Pitfalls and Practical Constraints in A/B Testing
Anti-patterns, A/A tests, real-world challenges (Ch.2.7, 3.1–3.2)
Week 6 – Improving Experiments
Flexible designs, covariate control, heterogeneous effects (Ch.3.3–3.5)
Week 7 – Introduction to Difference-in-Differences (DID)
Basic concept and intuition (Ch.4.1)
Week 8 – Practical DID Analysis
Applications, multi-period DID, parallel trends (Ch.4.2–4.4)
Week 9 – Regression Discontinuity Design (RDD)
When RDD applies and key assumptions (Ch.5.1–5.2)
Week 10 – Practical Issues & Student Presentations
Method selection in practice, limits of analysis, project discussion (Ch.6) |
|
|
履修条件・関連する科目
Course Prerequisites and Related Courses | | | There are no formal prerequisites; however, basic knowledge of Python programming is required. |
|
|
成績評価の方法と基準
Course Evaluation Method and Criteria | | [Course evaluation]
1. Participation & Attendance – 70%
(1) Active participation in class activities is essential in this course. Activities include in-class exercises and discussions.
(2) Attendance will be recorded at every class meeting. Students who miss more than two classes without a valid reason will receive a 20% deduction from their final course grade.
2. Homework – 30%
Homework assignments will be given after each Python unit.
[Grading Schema]
To pass this course, you must receive a C- or higher grade. The grading schema is the following:
Score Grade GPA
95 - 100 A+ 4.3
80 - 94.99 A 4.0
70 – 79.99 B 3.0
65 – 69.99 C 2.0
60 – 64.99 C- 1.0
0 – 59.99 F 0 |
|
|
教科書・参考書
Textbook/Reference Book | | | Yasui, Shota. Pythonで学ぶ効果検証入門 (Introduction to Impact Evaluation with Python). [in Japanese] |
|
|
課外学習等（授業時間外学習の指示）
Study Load(Self-directed Learning Outside Course Hours) | | | No advance preparation is required. However, students are expected to review the material after each seminar and complete the assigned homework. |
|
|
注意事項
Notice for Students | | | Students must bring a laptop computer to every class for Python exercises. Smartphones and tablets are not suitable for this purpose. |
|
|
授業開講形態等
Lecture format, etc. | | |
|
遠隔授業（オンデマンド型）で行う場合の追加措置
Additional measures for remote class (on-demand class) | | |
|
質問への対応方法
Office hour | | | Students can come to my office during office hours or email me at okajima.hiroko.b7@f.mail.nagoya-u.ac.jp. Office hour information will be provided during the first class meeting. |
|
|