Divergence between Human State Assumption and Actual Aircraft System State

One of my Graduate Resident Tutors (GRTs), Sathya Silva, gave her thesis defense yesterday on the topic of divergence between human assumptions and actual state in aircraft. In short, the divergence framework she built examines how differences between pilots’ mental models of the state of the aircraft they are flying (i.e. throttle setting, landing gear position, etc.) and the actual state contribute to accidents and postulates mitigations for these divergence events. Her presentation (slides here) drew on a variety of sources, including previous research and aircraft incident reports. I’d like to focus on slide 33 in particular:


This slide comes after more detailed analysis of a number of incidents during flight and shows visually timing information for each incident in the divergence framework. The use of lines that split and come together again, with visual markers for recovery time, loss of control time, and impact times, highlight the extent of divergence in each case, any subsequent re-convergence, and the relative timing in events. One of the main takeaways from this aggregate view is that in many cases the difference between fatal and non-fatal incidents was not that convergence did not happen, but that it it happened without enough time left for a full recovery, proving insight into a possible mitigation strategy.

The audience for this defense was her thesis committee, as well as friends and family attending the defense. In addition to Sathya’s main goal of passing the defense, the presentation was meant to lay out the concepts behind divergence to provide evidence for the thesis; slide 33 in particular highlights patterns in these incidents in the context of the divergence framework. The presentation overall was effective in achieving its goals; the novel divergence visualizations provide at-a-glance comprehension in the divergence framework and demonstrate its effectiveness in highlight patterns. The thesis committee agreed that this was a great presentation – congratulations Dr. Silva on passing your thesis defense!

Why do we have more boys than girls?

Why do we have more boys than girls? is a YouTube video created by Minute Physics. The general audience for this YouTube channel is curious internet browsers who have a few minutes on their hands to learn a little bit more about how the world works. In this particular video, they explain why there are more boys born than girls.

One of the data points they explore is the genetic odds that a child is born as a boy or girl. This is demonstrated in the screenshot captured below. They depict a mother and father tossing each tossing a coin with an X or Y chromosome into the air. The coins meet and combine into either XX or XY and fill their respective sides of the scale with either male or female babies.

Screenshot (3)

The use of coins emphasizes the point that biologically, there is a 50% chance of a parent donating one of their two chromosomes. The scale further drives home the point that at the end of the process, there should be an equal ratio of boys to girls.

The general style of the video is a simple, well described animation that walks the viewer through a “proof” of how the theoretical ratio of 1:1 boys to girls being born is not true. This method is very well tailored to the target audience as it does not assume much technical knowledge and does not take more than 3 minutes to explain.

Paul Ryan vs. other Speakers of the House

Explore the visualization here!

Screen Shot 2016-02-03 at 7.15.25 PM

Bloomberg Politics put out a beautiful data visualization in light of the recent election of Paul Ryan to Speaker of the House. While their chosen title highlights the recent trend of Speakers no longer working for the government after the gig, there is actually a treasure trove of information to explore here. All past Speakers are listed, and their involvement in various government positions are represented in a timeline. There are also several options at the top to highlight each position, with various callouts providing more facts.

Continue reading Paul Ryan vs. other Speakers of the House

Concussion Awareness Poster

Heads Up Concussion

Concussion graphic

This poster provides information about the danger of concussion in sports. It is organized into a number of different rectangles, and each rectangle conveys a different point about concussion, with a loose progression from the top to the bottom. The goal of this poster is to keep kids safe during sports by educating people who may be in contact with them, but may be unfamiliar with what to do when a head injury occurs.  In most cases this likely includes parents and coaches (assuming that athletic trainers or school nurses have already been educated).

This poster attempts to accomplish its goal by starting off with a scary statistic about head injuries in kids, and then proceeding to give information about how to help kids avoid becoming a statistic. It urges adults to have kids wear helmets, and also provides information about concussion symptoms, and what to do if a kid sustains a head injury. I like the simplicity of the graphics, and the way it doesn’t provide an overwhelming amount of information, which makes the information it does provide more accessible, and thus more powerful. However, in some places I found the information a little too simple, to the point where it became overly vague. For example, the “scary statistic” mentions TBI, not concussion, which is in the title, yet fails to define TBI as traumatic brain injury. It is unlikely that adults who are learning about concussion for the first time would be familiar with the acronym TBI.  It also does not provide information about when an athlete should not only be removed from play, but should also seek immediate emergency help, which seems like an important oversight, and it doesn’t indicate what might happen if a player returns to play with a head injury (which has serious health consequences). Overall, I think this poster does provide a really basic education concussion, and keeps things simple, which is important for audience comprehension, but needs to provide a little more information in some places to truly be successful.


Enrollment Request

Due to the design of the course I have to limit enrollment to 25 people. I prioritize enrollment based on seniority, affiliation, and background. So if you are a grad student / upperclassman, from MIT/Media Lab/CMS, and bring a unique skill among the student set, then you are more likely to obtain admission.

Fill in this short enrollment survey to let me know you are interested in enrolling and why.  I will get back to you before Thursday on whether you are in or not.

UPDATE: All emails have been sent.  email rahulb [at] mit [dot] edu if you didn’t get an email!

Course Description

We are swimming in data – “Big” and small, global and personal. And we are also facing complicated problems like Climate Change and inequality whose stories can only be told with data. The need for public understanding of data-driven issues is higher than ever before. But raw data doesn’t make a good story and that’s where you come in.

This class is focused on how to tell stories with data to create social change. We will learn through case studies, examples and hands-on experimentation with tools and technologies. We will teach basic methods for research, cleaning and analysis of datasets, but the focus in on creative methods of data presentation and storytelling. We will consider the emotional, aesthetic and practical effects of different presentation methods as well as how to develop metrics for assessing impact. For the final project, students will create a piece that tells a data-driven story.

The course is open to all technical levels and backgrounds. We will prioritize students with a strong background in one or more of the following areas: journalism, software development, data analysis, documentary, visual and performing arts.

This semester the course will have a special focus on Civic Data – using data for the public good. Most examples will use data related to this topic, homeworks will be related to it, and final projects must be connected to it as well.

We will meet Tuesday and Thursday from 3 to 4:30pm in MIT room 4-261.

Read the full syllabus for more info.


This is the shared class blog for the CMS.631 / CMS.831 Data Storytelling Studio course at MIT (Spring 2016).  Many of your homework assignments will require you to submit blog posts here.  Feel free to cross-post them to your own blog.  Much of the conversation about data, finding stories, creating presentations, and creating change happens online; you need to add your voice to that conversation if you plan to do this type of work.