15.276  Communicating with Data

Please bring Laptop to class

Please bring your computers to class today. We will set aside some time for you to complete the course review. Since we have so few people in the class, and this is the first time the course as been offered, we would very much like to shoot for a 100% completion rate.

Announced on 10 May 2016  10:38  a.m. by Miro Kazakoff

Your final presentations

We are one week from your final presentations for the class (this is actually your 2nd to last assignment as your final assignment will be a self-reflection due on May 12th). Please remember to familiarize yourself with the details of the assignment as you plan your presentations.

25-Minute Team Presentation of Findings (20%)

Based on the data analysis your team has done individually, each team will create, rehearse and deliver a 20-minute presentation with an additional five minutes of Q&A and one minute of context setting before the presentation. The presentation will be strictly timed and every student in the team must present.

As with the write-ups, you must specify a specific audience for the presentation. This audience ought to be similar to the one for your write-up (e.g. if your write-up was to one consulting partner, this presentation would be to a group of partners). In your one minute to set context, it should be clear why this group has invited you to present and what the goal of the presentation is.

Condensing the findings of five people into 20 minutes is intended to present a substantial, but realistic challenge. We would rather see you explore a few points deeply than provide a thin overview of everyone’s work. Not every student is expected to present their analysis (though every student is expected to present something. Good presentations will:

• Tell us why this work is important. In most cases it will be appropriate to extend your results into a recommended course of action.
• Intelligently choose what data to focus on. Even though everyone must present, you do not have to share findings from every team member
• Create a structure that presents your findings in a logical flow
• Share your structure with the audience to aid in understanding
• Have a clear introduction and conclusion
• Answer the “so-what” question of why this analysis is meaningful and worth having done
• Transition seamlessly between presenters
• Use visuals to convey information that can best be conveyed visually (rather than act as a script for the presenter)
• Use best practices in visuals that we have discussed
• Have been rehearsed multiple times before class
• Handle Q&A effectively as a team

Announced on 28 April 2016  3:27  p.m. by Miro Kazakoff

Pre-reading for our speaker tomorrow

Class-

We will have a virtual guest speaker tomorrow from Uber. He is going to talk us through a real life project he did, so please take the time to look through the materials below. You don't need to have them memorized, but please take 10-15 minutes to familiarize yourself with them so that we can have a discussion respectful of the time he is giving us.

To Read

1. The four links on his research page under CensusMatchR: http://steven.ae/research.html
2. His one sheeter post: http://steven.ae/onesheeter_story.html
3. Hal Varian on the importance of Data Science: https://www.youtube.com/watch?v=D4FQsYTbLoI

See you all tomorrow.

Announced on 25 April 2016  10:08  a.m. by Miro Kazakoff

Presentations on May 5th

Hi All,

We will be providing lunch for the class presentations on May 5th. Please email me (amyruth@mit.edu) no later than Friday, April 29th if you have dietary restrictions (vegetarian, vegan, gluten-free, allergies).

Thank you,

Amy Wasserman

Announced on 22 April 2016  8:52  a.m. by Amy Wasserman

Data Impromptu Sign-ups

In preparation for our data impromptu class, please provide a link in the Google Sheet below to the article or data source you want to speak on. Your selection is due by class time Thursday 4/28.

You may choose the same article as someone else. You can imagine some speaking scenarios this is likely to generate. You can decide for yourself if that is an argument for or against it.

Again, the article must have a chart of graph associated with it or a substantial amount of quantitative data. We will expect you to use numbers in your impromptus.

A few sites I reliably find interesting for these things:
http://fivethirtyeight.com/
http://www.economist.com/
http://polygraph.cool/
http://www.nytimes.com/section/upshot

Please enter your choice into the google sheet linked here:
https://docs.google.com/spreadsheets/d/14u8BTGVA36hH7I2M8tMf5SAfNerRayJLKdoLnwwNp1M/edit?usp=sharing

Announced on 21 April 2016  6:34  p.m. by Miro Kazakoff