Design Exercise 10: Project Proposal

in Design Exercises
Page content

For the last part of the class, we will have everyone do a “project” based on the ATUS data that we are now all familiar with. The first phase of the project is to propose a project (to be turned in as DE10 (due Tue, Nov 22)). There is actually a part before that: in class on November 16, we will have a class discussion, brainstorming session, and partner finding exercise, which means you must prepare for it by reading the project documentation (this page and some others) and completing DE10A: Project Topics Survey (due Tue, Nov 15).

Warning: the time frame for this project is very compressed. We will adjust our expectations accordingly.

While this assignment is officially released, expect updates and clarifications as we progress.

Overview

The project proposal is the first phase of a larger project. This page serves as both the instructions for the proposal, but also the documentation for the larger, overall project. You need to know what the overall project will be, so you will know what you will be doing so that you can make a decent proposal.

The project involves the ATUS: American Time Usage Survey data set that you have been working with. However, it is a bit more open ended. We have chosen three Project Themes - you must pick one. But within each theme, there is a lot of flexibility. You might build an interactive tool, try to design a specific visualization for a hard task, explore the space of possible designs, etc.

The proposal itself is a modest form: DE10 (due Tue, Nov 22). It will ask you which theme, and for a brief description of your intents. It will give you the opportunity to provide more details about your plans (if you know them). The more you can tell us about your plans, the more we can help you. However, the main thing we want to do is to check that you really have thought about the project.

The final project will be due the last week of class: DE13: Final Project Handin (due Fri, Dec 16). The due date listed on Canvas is the Tuesday (to keep the symmetry with other Design Exercises), but, in practice, we will allow you to turn in your project a few days late without penalty. There will be a “drop dead date”: if you don’t give us enough time to grade things before we have to get grades turned in, we can’t give you a grade.

There will be intermediate check-ins so we can assess your progress.

Project Details

Note, this may be updated as more aspects of the project are clarified.

Project Themes

For the project, you must pick from one of 4 different themes. The themes are discussed at Project Themes. Each theme is broad enough to allow for some creativity, and for different “styles” of project. But hopefully, the themes are chosen so you can do something reasonable in the short time we have for these projects.

Your project must be responsive to the theme you select. If you do not address the problem described in the theme, you will not get a good grade.

You will need to select a theme as part of the project proposal. If you want to change your theme, you must discuss it with the Professor (send a note on Piazza) and get permission.

Working Together

We will allow (encourage) students to work with a partner. We will have an in class exercise to help students meet potential project partners.

We will allow groups of 3 to work together, but these must be approved by the professor before the proposal deadline. If you want to work in a group of 3, contact the professor (via Piazza private message) with a brief note saying (1) who the partners are, (2) a brief description of the project ideas, and (3) an idea of how you intend to divide the work among 3 people. Please submit this at least 24 hours before the project deadline so that we can review it. (we expect to approve all requests where at least some thought has been given to how to divide the work among 3 people).

A problem with partner projects is that they depend on who you know (can you identify a good partner?). We will try to mitigate this problem by having opportunities to help students identify partners.

A partner project must be a true partnership. It is usually clear when it is two projects that have been stapled together. Teams will turn in one final result (each student will turn in a self reflection independently), so there is less preparation work (e.g., only one report and video), but we will have higher expectations.

If students choose to work together: (1) both must agree; (2) both must submit the proposal independently (but explain how they will work together) - there will be an opportunity on the proposal form to indicate ; (3) the team may not split up after the proposal phase without permission from the Professor; (4) both teammates must do the individual aspects of the assignment (e.g., self assessments, evaluation exercise, peer feedback, …), (5) except in extreme circumstances, the project grade is assigned to the team.

Note: team projects must really be team efforts. Having different team members do separate things that are just stapled together at the end is not a team effort.

Types of Project

This project will require you to work with the actual data. You’ve already done some of it - now you can do more. If you want to work on a subset of the data (for example, randomly downsample, or only use select years), you can - but be clear that you are doing so. Making things that work on the entire huge data sets adds to the challenge.

Projects may be more “implementation oriented” (where you build things and describe them), “design oriented” (where you focus on coming up with ideas and assessing them without really building them), or analysis oriented (where the goal is less to make a tool, but rather to show visualizations that are the result of some analysis). Many projects will mix these elements.

If a project is a Design Exploration, it will be important to come up with multiple ideas and compare and contrast them. In this case, the fidelity of the individual design doesn’t need to be as good: you might make a number of static designs, or even sketches. It is unlikely you can build multiple interactive approaches - unless you are very good at prototyping.

If a project is more implementation / systems oriented, it will be important to demonstrate how your tool is useful on the real data set. Note that we may not be able to run your tool (unless you are able to host it on the web) - so documenting its usefulness is part of creating it.

If a project is more analysis oriented, it will be important to both find interesting patterns in the data, as well as presenting it in a way that shows the results in an interesting way. The interesting part should focus on how the “result” is shown, not the methods used to find the “result.”

What to expect to turn in at the end

The details of the final handins will be provided closer to the end. But I want you to have a sense of what you are going to produce so you can plan accordingly.

Project Overview Documents:

  • Report: This will be a PDF document, in single column format (with a reasonable font, …). The target is 4 pages of text - plus images and references. This is inexact (we prefer that images are integrated), so its hard to just count text pages. We do value clarity and conciseness in exposition. If you find yourself needing a lot more than 4 pages of text, consider only making the key points and putting extra information in a supplement. Remember, we need to grade a lot of projects in a very short amount of time.
  • Video: You must provide a 2-5 minute video giving a summary of your project. 6 minutes is a hard limit (if your video is longer than 5 minutes, it will be held to a higher standard). It is OK for this to be a recorded presentation (you talking over powerpoint), or a mix of slides and demos.
  • Self-Evaluation: You will submit a self-evaluation on a Canvas form. We will give you specific questions to address. If you work with a team, you will be asked to discuss how the team worked together.
  • Description: we will ask you to describe the artifacts that you turn in, what we need to do to look at your project, and a brief “abstract”. If you have programs, you need to give us details on how to run them.

Other handins (as appropriate):

  • Code artifacts: you may either turn in a zip file, or invite us to view a GitHub repository.
  • Demos: please try to make any systems/demos run on the web (so you can provide us with a URL). It is unlikely that we will try things that require installation, or doing things to run locally. We have a bias towards projects that we can play with online.
  • Extra images and supplementary data. If things don’t fit into the project report, you can put them here. If you find yourself needing to give lots and lots of images, you might not be thinking hard enough about how to use visualization to summarize your results.

Timeline

Note: these links are to Canvas, which is how things are turned in - not to the course web pages that will describe them.

The Project Proposal

The actual project proposal (turned in on Canvas DE10: Project Proposals (due Tue, Nov 22)) will ask you:

  1. Which theme are you doing
  2. Are you working with a partner (and if so, whom) - note everyone must turn in a proposal
  3. A brief description of your intended project - this should describe the problem you intend to solve, and the strategy; not necessarily all the details of the ideas. For example, you might describe that you are going to try to devise an interactive system for exploring X, not necessarily the specifics of the design. This should convey a sense of the expected result.
  4. A brief description of the intended “implementation” - what tools do you intend to use? This is mainly for implementation projects (list languages, platforms, …), but even paper design projects should discuss this (e.g., will you use tools to help you explore the data, etc.)
  5. (optional) More detailed ideas. We understand that you might not have figured out how to do things - but if you have ideas (something you want to try, …) we encourage you to describe it.

We expect the proposals to be brief, but the more you tell us, the more likely we are to be able to give you useful feedback.