Submission instructions

Submit a PDF (1 page or more) on Gradescope. One submission per team.

Overview

The course project is an open-ended data analysis. Two requirements distinguish it from the mini-projects:

• You must collect your own dataset. Re-using an existing curated dataset (Kaggle, UCI, course repositories, etc.) is not permitted. Acceptable sources include scraping, surveys, public APIs, sensor logs, or merging public sources in a way that produces a genuinely new dataset.
• The deliverable includes an interactive dashboard (e.g., Shiny) in addition to the written report.

The proposal is a short document that lets us check that the question is well-posed, the data can actually be collected, and the planned methods are reasonable. You will receive feedback before starting the project in earnest.

Section 1: Team and topic

State the team members and a one-sentence description of the project topic.

Section 2: Question

State the specific question your project will answer. A good question is narrow enough to answer with the data you have, but substantive enough that the answer is not obvious in advance.

Section 3: Data

Describe the dataset you will collect:

1. The source(s) and the collection method (scraping, API, survey, sensors, merging of public records, etc.).

2. Roughly how many observations and variables you expect, and what each row will represent.

3. Any access, licensing, terms-of-service, or ethics considerations.

4. A short justification of why the resulting dataset is unique — i.e., not already available in curated form.

Section 4: Methods

Sketch the planned analysis. Name specific techniques (e.g., logistic regression, hierarchical model, bootstrap). If a model is involved, write its form, e.g.

\[ Y^{(i)} = \beta_0 + \beta_1 X^{(i)} + \epsilon_i, \qquad \epsilon_i \sim \mathcal{N}(0, \sigma^2). \]

Explain in one or two sentences why this method matches the question.

Note: there would not necessarily be a single “right” model, or a model at all. The point here is to start thinking conceretely about the project methods. This will not be graded harshly.

Section 5: Plan

A short timeline with milestones and who is responsible for each. Two or three rows is enough.

Report quality

The proposal is graded on clarity. In particular:

• The question is stated precisely, in one or two sentences.
• The data collection plan is concrete and feasible within the term.
• The proposed methods are named, not just gestured at.