Project Description

Overview

You will do a survey study on climate change perceptions. You’ll collect responses from N people (N being... however many you can guilt into doing it), then poke around (i.e., apply Data Science) looking for links, for example, between their personalities and what they reckon about the climate. Earth's dying - let’s get some charts about it, shall we?

Deliverables? Of course:

A report (maximum 8 pages for main paper, because my attention span is like a goldfish’s)
A link to a downloadable video presentation (yeah, it will be fun!). This should ideally be a 3-minute video pitch of your project (like Nima's. He and his business partner are Master students and won Slush 2024. If they did it, you can do it). More realistically, the "minimum viable video" is a video showing a powerpoint presentation for 3 minutes.

Deliverables Deadline: July 3rd.

Survey Design & Data Collection (How to Annoy Strangers with Forms)

For each participant of your N, you must collect data on the following:

Climate Survey (Italian Version by ChatGPT) – What people say they care about vs. what they actually do. Classic.
Big Five Personality Test (Italian Version by ChatGPT) – Because who wouldn’t want to know if they’re a neurotic introvert saving the planet? This is how you score the Big5.
An Additional Survey (Italian Version by ChatGPT) – Made it for you to make your life more exciting.
Personal Characteristics – Your survey participant's gender, country of origin, age, zipcode (to consider their area's media income) , and any chracteristic you decide (surprise me!). NOT names or emails. Bonus tip: You can give them a random ID so you know who’s who without actually knowing who’s who. GDPR and all that jazz.
One question (or very short survey) of your choice. Have a creative moment. In my "moment", I thought about asking the Moral Foundation Questionnaire (and this is how you score MFQ. This is the Italian MFQ provided by your best friend, ChatGPT).
Signed Consent. Each participant needs to sign a consent form. If a minor, a parent needs to sign a parental consent form. All signed consent forms need to be attached in the Appendix of your report.

Data Analysis Tasks (Let’s Pretend We’re Scientists)

You’re going to do some “analysis”. Sounds fancy, but it’s pretty basic:

Correlation Matrix – Draw a pretty table showing which questions are connected to what (ALL questions).
Stratified Analysis – Split your data by personal stuff like gender or country and answer these:
- Do men and women answer differently?
- Which questions caused the biggest “aha!” moments?
- Any obvious clusters, like a bunch of anxious vegans in one corner?
Predictive Analysis – Basic requirement: here you must use both linear regression and logistic regression, as requested in Homework 4. Beyond that: Get creative. Use any classification/predictive algorithm you like. Can certain answers predict if someone’s hopeful or just... crying about the future?

Grading Criteria

Here’s what matters to whoever is marking this:

(10%) Your sample size (your N): More people = better. More specifically: (N / (#team members-7)), where min(#team members)=8.
(10%) Class’s N: Total sample size from the class. So yes, your classmates’ laziness will drag you down.
(10%) Your team diversity: Different skills, backgrounds, countries. Basically, if everyone’s called “Antonio” and studies Environmental Engineering, rethink your team.
(5%) Quality of your first video. This is the video you have created as part of the class project warmup assignement.
(15%) Quality of your final video. This should ideally be like a pitch video of your project.
(20%) Quality of research questions and how well you waffle about them in your report and video.
(30%) Quality of your report. This accounts for the majority of your grade because it is extremely important that your report looks as polished and professional as the paper you have read.

Incorporating Peer Evaluations into Grades. This is the average score assigned by your team members shortly after submitting your report and video (by "shortly after", I mean within the same day). Peer evaluations are commonly used to assess individual contributions and performance within a group. These evaluations are typically anonymous and designed to provide constructive feedback, promoting accountability and supporting a more holistic learning experience. Therefore, your final project grade will be 80% group grade (that above) + 20% average peer score (scaled to 100). Also, I will adjust grades manually based on narrative comments and score patterns from your team mates. Anyone who does not submit a peer evaluation for all their teammates will automatically receive a score of 0 for his/her average peer evaluation.

Zero Grade Policy (Brutal but fair):

Grade = 0 if:

You make up the data? Zero.
You forget to submit? Zero.
You’re late? Zero.
You submit not according to instructions (e.g., a report of 9 pages as main paper)

Report Structure (Max 8 Pages for Main Paper)

Here’s the structure. Stick to it. In addition to title and authors' names, the report should contain 6 Sections:

Abstract – 3-4 sentences. Situation, Complication, Proposal.
Data Collection – How you annoyed people into answering your survey. How many did you collect? Who are they (e.g., age, gender)?
Research Questions – 2 to 4 clear, actually interesting questions.
Methods – How you tortured the data, aka how you analyzed the data.
Results – Answers to your research questions, with interpretations plus your big “Aha!” moment.
Appendix (outside the 8-page restriction) – It should contain:
- Declaration – If you used ChatGPT, just say it.
- Division of Labour – Who did what. Don’t lie.
- Team Diversity – Brag about how “global” and “interdisciplinary” you are.
Writing Tip – Do a pyramid for your entire report. It will help you a lot to write your sections and subsections.

Final Submission

Turn in:

The report (PDF, max 8 pages). Submissions must be in English, in double-column format, and must adhere to the ACM template and format (also available in Overleaf). Word users may use the Word Interim Template. Submissions must be a single PDF file: 8 (eight) pages as main paper, with unlimited pages for references and an optional Appendix (that can contain details on reproducibility, proofs, pseudo-code). The first 8 pages should be self-contained, since we are NOT required to read past that to grade your report.
A link to a downloadable Video Presentation specified inside your PDF report.

Final Notes (Read this. Or don’t. Up to you)

Pyramids, Pyramids: Use the Pyramid Principle to write your sections.
Stay focused: No rambling about capitalism unless backed by your stats.
Formatting Requirements: 8 pages. Not 9. Not 47.
Make your research questions compelling: “Do people care about the planet, or are they just tweeting?” That sort of thing.

Final Final Notes (mainly about translating the Survey in a language other than English)

Some questions were rephrased:
- Questions 4, 9, 11, 28 – "negli Stati Uniti" in the initial Italian translation to "in questo paese"
- Questions 21, 26, 27 – changed "Presidente" in the initial Italian translation to "Presidente del Consiglio"
- Questions 22, 27 – changed "Congresso" in the initial Italian translation to "Parlamento"
In the literature, previous analysis recoded answers into binary (0/1) outcomes:
- “Yes” or “Strongly agree” = 1 (positive)
- “No”, “Disagree”, or “Don’t know” = 0
This recoding (if you do any) should be detailed in the Methods section of your report.

FAQ

Do we have to use the survey you provided, or can we modify it? You must use the survey questions I provided exactly as written, as they are standardized. However, you are welcome to add additional questions or items if you'd like.
Do we need to provide the surveys on separate pages? Not necessarily. You can administer the surveys all at once or in any format you prefer, as long as the questions are the standardized ones.