MSDA Final Project Guidelines

Submission guideline

You will submit:

  • All code (not applicable if you used only Tableau)
  • Datasets
  • Slides in Powerpoint, Keynote, or HTML (if used)
  • Tableau workbook (if used)

All the above files should be zipped and submitted on Blackboard before 10:30 pm on 10/13 (i.e., the deadline is 2 days after the last class).

The Zip file will have to be named as follows if you are from the day cohort:

DA6233_2018_Day_Group##.zip

Here Group## will be your group number. For example, Group number 1 will submit

DA6233_2018_Day_Group01.zip

Note the leading zero in the group number

The Zip file will have to be named as follows if you are from the evening cohort:

DA6233_2018_Evening_Group##.zip

Here Group## will be your group number. For example, Group number 1 will submit

DA6233_2018_Evening_Group01.zip

Note the leading zero in the group number

Grading guidelines for the project

Every group gets 20 minutes for the presentation. Out of these, 15 minutes are for the actual presentation and 5 minutes for Q&A. We will strictly stick to this time limit. Q&A may go beyond 5 minutes in case the audience have a lot of questions. But the presentation may not exceed 15 minutes.

In order to stick to your 15 minutes, please rehearse your presentation with all the group mates at least once.

The project will be graded on the following 4 criteria:

Are the insights interesting? (40% weight)

Interesting insights are defined as the insights that: - Are non intuitive

  • Might change knowledge and/or behavior of a lot of people at least marginally OR might change knowledge and/or behavior of a few people substantially

  • Are actionable

Evidence of hard work on insight discovery (20% weight)

This is somewhat related to the first point because unless you do a lot of exploration, it’s difficult to find interesting insights. However, a lot of exploration may not lead to interesting insights. In case your insights are not interesting, I will definitely give you credit if you provide evidence of significant exploration. This could be in the form of all the plots or Shiny apps you made. It could also be in the form of Tableau dashboards you built during the exploration stage.

Quality of the visualization (20% weight)

Good visualizations will have several properties that we have been learning during the course:

  • Easy to interpret. There is one message per visualization.

  • Easy to read. No clutter, no small fonts, no obscure fonts, right color palettes to increase the contrast, etc.

  • Right type of visualization. For example, if you have zip codes, use spatial visualization such as maps.

  • Interactivity only when necessary. A lot of times interactive visualizations may hide information by making the dashboard too complex for the stated purpose. Choose interactivity carefully.

Quality of the presentation (20% weight)

I will look for the following properties:

  • Did you stick to the time limit?

  • Was your presentation energetic? In other words, were you interested in your presentation?

  • Did you face the audience as opposed to looking at the screen (on into the abyss!)?

  • Did the transition from one group member to another went smoothly?

  • Did you handle the questions from the audience well?

Related

comments powered by Disqus