Assignments

This page describes each of the homework assignments, listed by the session in which they are due. It also provides some suggested datasets for each sketch, and selected examples of student work from the 2017 and 2016 classes on some assignments and the final project. (While the 2016 version of the class wasn’t specifically focused on climate change, the sample works are good illustrations of overall goals of the assignments.)

See the syllabus Course Requirements section for more details on these assignments and grading criteria.

SES #	HOMEWORK DESCRIPTIONS	SUGGESTED DATASETS & SAMPLE STUDENT WORK
Finding and telling stories with data
1	No homework due
2	Homework 1: Visualization blog Write a post on the class blog about a data presentation you saw recently (250ish words). Include a link and a picture/screenshot of the presentation. Make sure to address: what data is being shown who you think the audience is what you think the goals of the data presentation are whether you think it is effective or not and why
3	Homework 2: Data log Create an activity log of all the types of data you create and are captured in digital form during the course of one day, and enter it on the class blog.
4	Homework 3: Software installations Download and install the following: Sublime Text (or any text editor capable of running regular expressions) OpenRefine Trifacta Wrangler (free with registration) Microsoft Excel, or free LibreOffice Still if you prefer Import.io (open it to sign up) Jupyter (if you know how to program in Python)
5	Homework 4: Tableau installation Download the most recent version of Tableau Desktop for students
6	No homework due
Sketch 1: Charts and creative charts
7	No homework due	Suggested datasets: Hubway data challenges: 2011-2013 / 2017 NYC tree census data: 2015 / 2005 / 1995 World Bank CO2 emissions: global or per country (ZIP) (This file contains: 2 .csv files.) Or any other climate change-related dataset you are really excited about
8	Homework 5: Start Sketch 1 Come to class knowing what dataset you are working on, an outline of the story, and a sketch of what you want to make.
9	Homework 6: Sketch 1 finished, blog post Finish your sketch. Write up your project on the class blog. Start with a summary sentence in the form: “The data say ________. We want to tell this story because _____.” Include a 300ish word summary of your data sources and why what you made is an appropriate and effective way to tell the data story.
Sketch 2: Data sculptures
10	No homework due	Suggested datasets: IPCC global temperature projections (2007) from various models (using these emissions scenarios, global average temperature in Kelvin) Unofficial unverified air quality readings in Beijing from US State Department (judged via this standard scale) Or the Hubway, NYC Tree, or CO2 emissions data from Sketch 1
11	Homework 7: Start Sketch 2 Start working on your idea for this sketch. Come to class with with: The data you are going to use (remember to go data shopping!) The audience you want to make something for (remember “general public” is an unacceptable audience!) Good but not required: one abstract word in your data story that you need help representing in concrete ways (“inequality”, “hate”, “asylum”, etc.)
12	Homework 8: Sketch 2 finished, blog post Finish your sketch. Write up your project on the class blog. Start with a summary sentence in the form: “The data say ________. We want to tell this story because _____.” Include a sentence saying “Our audience is _______.” Include a 300ish word summary of your data sources and why what you made is an appropriate and effective way to tell the data story.	Sample student work: Data sculptures Fireworks: Fun and Dangerous (2016)
Sketch 3: Personal stories
13	Homework 9: How to tell a story (The Moth) Read Paige Williams, “How to tell a story: The Moth.” Nieman Storyboard (Harvard), August 2012. Choose and listen to one Moth Story. Write down and bring to class your analysis, as Paige Williams lists in her article’s item #3. Fill in the class midterm survey.	Suggested datasets: Pew Research Center. “Global Concern about Climate Change, Broad Support for Limiting Emissions.” 2015. Compiled public input to the GoBoston2030 campaign [_data not available for OCW_] Food for Free data on food scarcity and food rescue around Boston [_data not available for OCW_] Or any dataset from a previous sketch
14	Homework 10: Start Sketch 3 Start working on your idea for this sketch, and bring into class: data, audience AND goals!

15	Homework 11: Sketch 3 finished, blog post Finish your sketch. Write up your project on the class blog. Include a short title for your project, and all your team member names. Start with a summary sentence in the form: “The data say ________. We want to tell this story because _____.” Include sentences saying “Our audience is _______.” and “Our goals are _______.” Include a 300ish word summary of your data sources and why what you made is an appropriate and effective way to tell the data story.
Sketch 4: Participatory data games
16	No homework due	Suggested datasets: Hubway data: 2011-2013 Kyoto cherry blossom flowering dates: data and background NSIDC Arctic sea ice coverage (CSV or GeoTIFF files) Somerville MA tree data [_data not available for OCW_] NYC tree data US Fuel Economy measurements
17	Homework 12: Start Sketch 4 Start working on your idea for this sketch, and bring into class: data, audience AND goals!
18	Homework 13: Sketch 4 finished, blog post Finish your sketch. Write up your project on the class blog. Include a short title for your project, and all your team member names. Start with a summary sentence in the form: “The data say ________. We want to tell this story because _____.” Include sentences saying “Our audience is _______.” and “Our goals are _______.” Include a 300ish word summary of your data sources and why what you made is an appropriate and effective way to tell the data story.	Sample student work: Participatory data games Donate by Playing: A Fundraising Board Game for the International Federation of Red Cross and Red Crescent Societies (2016) How Much Do You Actually Know About Commuting In Boston? [Hubway experience quiz] (2016)
Sketch 5: Maps and creative maps
19	Homework 15: Map Somerville tree data Use Tableau or CartoDB to make a map of the Somerville MA tree data Watch the Tableau Basic Mapping Tutorial [9 min video] (requires Tableau registration) Online Mapping for Beginners tutorials from CartoDb	Suggested datasets: NSIDC Arctic sea ice coverage (CSV or GeoTIFF files) Somerville MA tree data [_data not available for OCW_] NYC tree data 2017 USDA bee colony data (PDF) by state 2015 NStar (now Eversource) gas leaks data, from HEET-MA Leak definitions: Grade 1 is in a contained space and so considered potentially explosive. Grade 2 is near a foundation and so must be watched. Grade 3 is everything else, no matter how big the total volume of emissions.
20	Homework 16: Start Sketch 5 Start working on your idea for this sketch, and bring into class: data, audience, goals, and question.
21	Homework 17: Sketch 5 finished, blog post Finish your sketch. Write up your project on the class blog. Include a short title for your project, and all your team member names. Start with a summary sentence in the form: “The data say ________. We want to tell this story because _____.” Include sentences saying “Our audience is _______.” and “Our goals are _______.” Include a 300ish word summary of your data sources and why what you made is an appropriate and effective way to tell the data story.
Final project studio
22	Homework 18: Start final project Read the final project requirements. Think about what topic, datasets, and output technique you want to explore for your final project. This could be building on one of the sketches you did before, or starting something new.
23	Homework 19: Begin group work Meet with your group to better define your final project. Come to class ready to tell me: the datasets you’re using, the audience(s) on which you are focused, the goals you have for that audience, and the main technique you want to use.
24	Homework 20: Testing and check-in Continue working on your final project. Be prepared to present what you’ve got so far to me. Remember that you need to try it out with a few people in your target audience to test if you are accomplishing your goals.
25	Final project completed and presentation Finish your final project. Post your methodology, impact, and presentation slides (if any) as a post on the class blog. Fill in the team-member assessment survey, one for each teammate. Prepare a 5 minute presentation of your final project. This could be us playing the thing you made, a video of it in action, photos from using/showing it somewhere else, or something else. Include an introduction with short descriptions of your datasets, goals, audience. Be sure to summarize your impact findings.	Sample student work: Final projects BeeSmart (2017) Save the Bees (2017) Affluent White Bostonians Unfairly Targeted by Stop and Frisk Practices…Mostly While Leaving SoulCycle (2016)

Screenshot of the project webpage.

By Catherine Caruso, Kendra Pierre-Louis, Tiffany Wang

Materials

Webpage article with clickable annotations

Methodology

This project builds on a previous project that combined a satirical Onion-style article with a fake map. Now, our piece appears on our own satirical website, The Olive, a nod to the real satirical website, The Onion. The satirical main text is annotated using Genius with real facts about stop and frisk practices in Boston. It also includes a map that toggles back and forth between the real data and the satirical data. At the bottom of the article we include an explicit call to action, and a number of different options for learning more and getting involved.

We wanted to see if satire could be used to both inform, i.e. like real news, as well as possibly affect behavior.

The main dataset used in this report is used the Stop and Frisk (FIO) dataset that the Boston Police Department released after being sued by the American Civil Liberties Union (ACLU). Catherine D’Ignazio provided us with a geocoded version of the dataset that we used to create maps. The data indicate that the areas with the highest density of stop and frisk incidents are also the areas with more people of color, and most people that are targeted by stop and frisks are black men. We also used US Census data to provide the community descriptors – median incomes, and racial demographics – of the communities mentioned in the article (both the satirical data and the real data), along with Catherine’s extensive knowledge of Gisele Bündchen and Tom Brady.

The website is hosted by GitHub and was created in Jekyll. The annotations were created in Genius, and draw from a combination of the ACLU report, Census data, and the actual stop and frisk dataset. The satirical article includes the same stop and frisk statistics, but flips them so they apply to affluent, white citizens in Brookline and Milton. It also cites general race and income information for those neighborhoods. To avoid copyrighting issues, we included a photo of Gisele that is available for public use and photos of friends who agreed to be photographed for this project.

For the maps, we used cartoDB’s density map function to show the density of stop and frisk incidents happening in the areas around Boston. The density map split the data into seven different bins. In cartoDB, we altered the code to change the color scheme (cartoDB only has one inverted color scheme, which is from white to pink). We also created a manipulated map, where we flipped the color scheme, and used PowerPoint to add stop and frisk incidents to the Brookline and Milton areas. We used JuxtaposeJS to create a map with a slider that moves between the actual and manipulated versions.

To assess the efficacy of the website, we did a series of five informational interviews with people. In all of the interviews we asked a pre-question:

“On a scale of 1-5, with 1 being the least, and 5 being the most, how much do you know about stop and frisk practices in Boston?”

If the respondent said three or higher, we followed up with the question:

“On the same scale of 1-5, how problematic are current stop and frisk practices in Boston?”

Then the interviewee would go through the website. We used a screen recording the website to see how the reader’s engaged with the website. Afterwards we asked them a series of informational questions:

How did this piece make you feel? (prompting questions if respondent asked for clarification: Did you think it was entertaining? Did you think it was surprising?)
What was your overall reaction to this piece?
How did the balance of humor and real facts work for you?
Did you notice the annotations and click on them while you were reading?
Could you tell the difference between the joke data and the real data?
Did you trust the real data?
Do you feel motivated to do something about this problem?
Which action at the bottom would you be most likely to take?
Did this article change your opinion of stop and frisk practices in Boston?

As a final information gathering mechanism, we installed Google Analytics on the site to track visitor behavior.

Google Analytics visitor data for the page.

Once the site was completed, we encouraged people in our network – via email, Facebook, and Twitter – with connections to the Boston area to read the site. Some of the tweets used bit.ly to further allow for additional engagement tracking.

Goals & Audience

According to the ACLU, current stop and frisk practices in Boston are highly problematic-they disproportionately target people of color in low income neighborhoods. In fact, many people in these neighborhoods are subjected to so many stop and frisks that their daily lives are disrupted. Rather, than replicate them here below, you can find the complete statistics about stop and frisk practices in Boston in the annotation layer of our article.

Unfortunately, the individuals that are subjected to unfair stop and frisk practices often struggle to have their opinions heard by those in positions of political power who can actually change these practices.

Our goal with this project is to educate more people about stop and frisk practices in Boston. Specifically, we want to target people who have not been subjected to stop and frisk practices, and have not experienced racial or socioeconomic discrimination from law enforcement firsthand. In Boston, this includes middle to upper class Bostonians, many of them white, who may recognize the term stop and frisk, but know very little about whether or not it is a problematic practice. Because of their affluence, race, and status, people in this demographic tend to be in a greater position of power to bring about political change that can improve the situation.

To accomplish this goal, we wrote a satirical article that blends a light hearted, humorous fake news story with the facts about stop and frisk practices in Boston. We chose to incorporate the facts into the article as an annotation layer, where the reader can access them as pop-ups while moving through the article. The map toggles back and forth between a manipulated version that matches the satirical content of the article (where most stop and frisk incidents occur in Brookline and Milton MA) as neighborhoods with high numbers of stop and frisk incidents), and the real data (where stop and frisk incidents are concentrated in Dorchester and Roxbury, MA). The article appears on The Olive, a website we built that is targeted at the middle to upper class, white Bostonians that we want to educate about stop and frisk.

We used humor in our article as a strategy for achieving a higher level of engagement with our audience, and making readers more open and receptive to learning about the many issues with stop and frisk practices in Boston. We wanted to draw in readers that might not choose to read a serious article about stop and frisk, but would be willing to learn about it within the context of satire, and are in a position to actually do something about it. Finally, in the call to action section, we also took advantage of the personal story: although we were unable to interview someone directly for the development of the site, we embedded a video the ACLU of Boston did that told the story of Ivan, a Boston resident who has been stopped more than 30 times.

We also offer readers a number of options for taking action on this issue, ranging from the simplest (sharing the article on Facebook or reading more about it) to more involved (donating to the ACLU or signing a petition to Mayor Marty Walsh). We hoped that by offering so many options for taking action, we would encourage readers to engage at whatever level suits them.

Project Assessment

To assess our project, we conducted five semi-structured interviews with undergraduate students at MIT (in an ideal world, we would interview people in our target demographic, but within the constraints of this project, we went with who was available). Our readers included two white students and three asian students. Overall, most of our readers recognized the term stop and frisk, and knew what it was, but did not know any statistics about it.

Several of our readers laughed out loud while making their way through the article, and enjoyed the humor of the satire. Others did not find the article humorous, particularly those that weren’t familiar with SoulCycle. This dichotomy of reactions emphasized that humor is highly subjective, and challenging to use effectively for the purpose of appealing to a broad audience. Again, interviews with people in our target demographic would give us a better sense of whether or not the humor works within our target context.

One technical issue we encountered is that our subjects did not realize the highlighted portions had pop-up statistics associated them, and read the article without clicking on them. With more time and technical resources, we would either make the statistics so they popped up when the mouse hovered over the text, or have the first statistic pop up automatically to make readers aware of the annotation layer.

It is always difficult to get readers to actually take action after reading an article, but our strategy of offering different options seemed to work-our readers were diverse in action they were willing to take action: two said they would sign a position, two said they would read more information, one said she would like the BLM Facebook group.

View video page

The BeeSmart final project video.

By Krithi Chandrakasan, Samuel Resnick, and two other MIT students. (Used with permission.)

Methodology

The USDA has been collecting data on the US honey bee population for the past several decades. It has been tracking the population of bees per state, the productivity per hive, and the price of honey. This data is available on the USDA.gov website in .txt form with a different page for each year. We cleaned it by grouping all the years together in one .csv file with columns according to: State, Colonies, YieldPColony, Production, StocksPricePound, ValueProduction, Year.

Upon first looking at the data there seemed to be random fluctuations in state’s honey bee population, except for one state, which consistently grew until it far surpassed all the others in honey productivity. This state is North Dakota. As soon as we saw this we became interested. Why did this state that is so rarely on our radar have such a significant growth in honey bee population relative to all the others? We started to do some additional research and began to see articles by beekeepers and conservationists saying that “North Dakota was the last best place to keep bees in America.” We found out that this is because the honey bee’s habitat has largely been destroyed across the rest of the country as prairies have been converted to farmland to product high yield cash crops. This loss of habitat due to farming and pesticide use has been given the name colony collapse disorder. Even North Dakota is not safe from this epidemic, however, and its own prairies are starting to shrink, seriously endangering the last of the American honey bee population. These prairies were once protected by the Conservation Reserve Program (CRP), which protected natural grassland across the midwest, but as farmers learned that they could make more money by planting crops, the CRP has been slowly pushed back. There are several organizations working to sponsor the planting of wildflowers instead of cash crops in farm fields as well as encourage advocacy at the national level. Pheasants Forever is one of these organizations that operates in North Dakota, and is encouraging conservationists to sign a petition to expand the CRP this coming year. This organization is intending to spread across six other states in the midwest in the next year to help preserve the honey bees habitat. This issue is not only a matter of habitat preservation. Since the honeybee is responsible for pollinating most of America’s crops, the security of our food source is at stake. This is an issue that affects every single person living in America today, and that is why we need to tell the bee’s story.

This story is about a large scale issue that affects the entire US population, across all states, but it has its roots at a small scale level. Because of this, we determined that our story lent itself well to a zooming in, zooming out approach – similar to The Powers of Ten by Ray and Charles Eames. Additionally, the story has a central geographic theme since it is about the role of a single state in the welfare of the entire country. We therefore decided to tell our story by starting with a wide lens looking at the US bee population changes over the past few decades, then zooming in on North Dakota as an anomaly. We zoomed into a specific town, with a specific apiary, owned by a specific beekeeper, and took a look at a single hive of bees as we explained his story. We then started to zoom out as we told the reasons for the delicacy of the bee’s situation and how grave the problem could become. As we approached our initial viewing point, we left the viewer with a positive message knowing that they could help by signing this petition. The viewer finds him/herself in the same position as they started in but now with a deeper knowledge of a complex problem that affects them directly and with a resource at their fingertips through which to help the cause.

Impact

There were a few short term goals we had for this data story. First and foremost, we wanted to educate our audience on bee colony collapse and provide them with a new perspective on the issue. In addition we wanted them to understand that bees are one of the largest factors that impact our food security. As an immediate action step we hope that this video will motivate them to buy honey. In the medium term we hope that viewers will sign the petition. Finally, in the long term we hope that viewers will become strong supporters of legislation that preserves habitat.

For this data story we target young Americans both in and out of the midwest who are interested in conservation and are active on social media. While narratives like this one are often targeted at the more liberal community our story aims to cut across party lines. The organization Pheasants Forever which we promote within the story has bipartisan support, as many conservative hunters support the organization because it helps conservation and allows them to hunt pheasants.

In order to evaluate the effectiveness and impact of our data story we showed the video to farmers market patrons, beekeepers, and a North Dakota resident. We framed the start of the video with questions asking what the viewer thought was the most important factor that affects food production and which state the most honey bees live in. As a follow up to the video we asked if their answer to the first question had changed, if they had learned anything new about honey bees, and if they would be more willing to sign the petition supporting CRP(detailed Q&A given below). We then gave viewers a small strip of paper with the link to the CRP petition which served as the call to action.

Based on the results of our user testing it was clear that our short term goals were met. Most of the participants had predicted crop disease was the most important factor for food production, however after watching the video all but one participant changed their answer to bee colony collapse. Additionally, after watching the video it was clear that they had gained new perspective on the plight and importance of the honey bees and learned where geographically the problem was taking place. We also met our medium term goals, as all participants seemed very excited to receive the slip with information and said they would sign the petition when they had a chance. Finally, based on our user testing it is too early to know if we met our long term goal, however the positive responses we received are a good sign that viewers will care more about legislation preserving habitat in the future.

Overall, we believe that our data story addresses the very important topic of bee colony collapse with a powerful narrative, and both educates and motivates viewers to take action to help solve the problem.

Q&A

Before questions:

1. Which of these factors do you think affects food production the most?

Drought
Bee colony collapse
Global population increase
Crop diseases
GMO

2. Where do you think most honeybees in the US live? (Select state)

After questions:

3. After watching this video, does your answer to question 1 change?

Drought
Bee colony collapse
Global population increase
Crop diseases
GMO

4. Did you learn something new about honeybees in America?

Yes
No

5. After watching this video, if someone asked you to sign the petition would you be more willing to do so?

Yes
No

Testing Responses

Person 1: Middle aged man, a tender at one of the stalls.

Crop Diseases.
Arizona
Same answers, still crop diseases.
Yes
Yes

Person 2: Young girl

Bee colony collapse
California
Same answer, bee colony collapse.
Yes, specially commenting that she didn’t know about North Dakota.
Yes, I would already have, but certainly would sign it after seeing the video.

Person 3: Young girl

Bee colony collapse
Texas
Same answer as 1.
Yes, commenting as well about the fact that they didn’t know about ND.
Yes.

Person 4: Young girl (Friends with Person 3)

Crop disease
(Where is somewhere warm…) Florida
Yes, bees.
Yes, commenting they didn’t know about ND.
Yes

Person 5: Man, middle aged

Crop disease
California
“Well, after seeing this video, clearly the bees are important”
Yes, commenting they didn’t know about ND.
Yes, however commented that “they doubt it would have any impact”

Image of game, cards and pieces on a tabletop.

The Donate by Playing game.

By Reem Alfaiz, Gary Burnett, Phillip D. Graham, and Argyro Nicolaou. (Used with permission.)

Materials

Presentation slides about the game (PDF)

Game board (JPG - 6.2MB)

Game cards (PDF - 1.2MB)

Writeup

The data say that there are many people immigrating to Europe. In 2015 alone, more than 1 million people arrived to Europe by sea. The numbers are increasing every year. In the first 3 months of 2016 the number of sea arrivals was 6 times greater, than the same time in the previous year. This influx of refugees will put a lot of stress on relief organizations, as they will now be even more limited in the number of resources they have available.

The goal of our project is to increase funding for these organizations via a game. Our audience are donors attending a fundraising event for the International Federation of Red Cross and Red Crescent Societies. These people are at the benefit because they want to donate and have some level of investment in the cause. Our goal is to encourage the attendees to donate by eliciting an emotional response via game play. Our game puts them in the position of refugees and shows them how the money they donate to IRFC will directly impact the lives of people trying to immigrate to Germany.

We want to tell this story to highlight the impact that relief aid can have on the life of a refugee. The journey to asylum can be painful and exhausting. Often times it leads to separation from one’s family, and sometimes it can even lead to death. We want to show that this is not how it has to be. There are organizations out there that provide relief and make the lives of refugees more bearable, and donating to organizations such as IRFC can have a direct impact on the lives of real people.

One of the most powerful data sources we used for our project were personal stories found online, that were documented by real life refugees. This helped a lot in the creation of our characters and what sorts of events can occur during an immigration across Europe. These personal stories both contribute to the accuracy of the journey and also help the players sympathize with situation and feel an emotional connection to the player.

Another useful data source was the UNHCR database of the popularity of various routes across Europe. These helped with the design of the gameboard. There are many paths the players could traverse, however we only decided to include those that were actually feasible. To achieve this we removed routes that included borders that were closed. We also chose to only include paths that many people have travelled across according to the dataset, as opposed to less popular options.

The rules of the game are simple. You are a refugee from Syria trying to get to Germany. Along the way you encounter various obstacles but also different kinds of help. Each player will be assigned a character. The characters are: Malika – a 26-year-old nurse from Aleppo, Adnan – a 10-year old boy from Latakia, Youssef – a 30-year-old man from Homs and the Alsouki family – a family of 4 from Damascus

All players start with 10 stamina points. You draw a card at every location, starting from the common starting point that is Syria. The card has 2 kinds of information on it: it tells you where to go next, and it also tells you how each transition affects your stamina points. Some cards give the player the option to purchase stamina points. These real-life donations all go towards the IFRC fundraising effort to help the national Red Cross and Red Crescent societies in Europe to deal with the unprecedented number of refugees and migrants arriving from the Middle East and Africa.

We want the players to empathize with the obstacles and the hardship that migration involves.

We want people to encounter obstacles in the game that will motivate them to donate small amounts of money that can make a big difference.

We did not want to create a competitive game – buying stamina points benefits everyone on the table, so to speak.

This course culminates with a final project that will integrate many of the skills and approaches we cover throughout the semester. The final project is NOT a sketch; you should be creating a real, functioning version of the idea you have in mind.

I anticipate that one of the small projects you do in the technique modules will grow into your final project, but that doesn’t have to be the case.

The final project must gather, analyze, and synthesize various civic datasets into a data-driven presentation. It must include:

Work produced as a team (ie. 3 or more students in the class)
- Your fellow team members will be providing feedback on how you contributed and that will impact your final project grade
A strong connection to climate change data — data for the public good
A compelling and creative presentation of your data story, based on the data, with a clear and well-defined audience and set of goals in mind
An accompanying co-authored “methodology” blog post describing how you found your data, cleaned it, analyzed it, and synthesized it into a story. This should include links to cleaned datasets you created online. This should be around 500 words.
An accompanying co-authored 500ish word “impact” blog post describing the following:
- A set of goals that explain why you want to tell them your data story
- A stated and well-defined audience for your data story
- A write-up of semi-structured interviews and/or pre-post surveys with a handful of people in your intended audience to gauge their reaction
A 5 minute presentation of the project (and if you create slides, post them to the blog as well)

Sample Final Project Ideas

Here are a few potential final projects, to provide a sense of scope and variety acceptable:

A long-form journalism article telling a data-driven story, with accompanying flat visuals, that tries to explain a complex issue of importance to a naïve audience
A physical sculpture for installation in a public venue that re-interprets data in an artistic way to attract attention and interest of potentially interested passers-by
An interactive online text-based game that leads users through a set of data to evoke an emotional response and a call to action to do something about the issue
A collaboration with a community group to analyze their data and create a set of visuals that help them tell their story to funders and community in a new way
An interactive installation suitable for a museum setting, presenting the data in a physical way using motors, Arduino, etc. to educate visitors about the issue
A comparative study of how people perceive and can operate on different presentations of data

Those are just a few directions you could go. This course is intended to encourage creative new techniques for telling compelling data-driven stories, and will showcase a variety of examples to spark novel ideas, so take these as examples rather than constraints.

Several modes of transportation - walking, bike share, personal cars, and a subway stop - in Harvard Square. (Courtesy of Connie on Flickr. License CC BY-NC-SA.)

By Catherine Caruso, Kendra Pierre-Louis, and two other MIT students. (Used with permission.)

Writeup

For this assignment we created a BuzzFeed quiz.

The data say that since Hubway’s inception in 2011, ridership has increased. In fact, earlier this month [April 2016], Hubway had its 4-millionth ride. However most of its riders were born between 1981-1986. People in their early to mid-thirties make up 38-percent of Hubway riders. Yet those born after 1991, currently only make up four percent of Hubway ridership. Getting young people to start riding Hubway is important for the system’s longevity because once people habits become entrenched it is difficult to get them to switch. This is why marketers target people during life transitions – marriage, pregnancy, etc – because it’s easier to get them to try something new while they’re already in transition. Similarly, we wanted to target young people as they made the transition from students to adults.

We wanted to tell this story because Hubway provides a vital service that complements mass transit systems, while reducing the carbon emissions associated with driving and improving the health and wellbeing of riders. In addition, Hubway doesn’t just benefit Hubway – it makes the streets safer for cycling generally and by proxy for pedestrians. Writes Emily Badger in a 2014 Washington Post article about the rollout of New York City’s bike sharing program, Citibike.

As more people bike and walk, cycling and pedestrian fatalities actually decline. That’s because the more people bike and walk, the more drivers become attuned to their presence (either on sidewalks or road shoulders), and the more cities are likely to invest in the kind of infrastructure explicitly meant to protect them (all of which further encourages more cyclists and pedestrians).

Our audience is 22-30 year old young professionals who live in the Boston area and need to commute to work. Our goal is to get more car and subway commuters using Hubway. Because it’s notoriously difficult to get riders to change habits, rather than an aggressive hard sell we thought we could get younger people to try Hubway through the game approach of a BuzzFeed quiz. We choose BuzzFeed specifically because according to its ad sheet:

It has 200+ Million monthly uniques visitors 50% are 18-34 years old
BuzzFeed’s in-house experts help the right audience discover a brand’s content across BuzzFeed and on social, specifically allowing us to target people within hubspots footprint through IP filtering
Brands can track content performance in real-time using BuzzFeed’s social dashboard, allowing us to assess performance, adjust content etc on the fly.

Our goals are to increase ridership among our specified market by 10% over YTD performance. For this experiment which has some limitations – namely that we’re not actually Hubway, that we’re using BuzzFeed’s community sharing profile as opposed to its viral advertising and don’t actually have BuzzFeed support, we’d like to have 500 views/clicks over the week in which we launched the quiz with a 50% completion rate. To date we have a total of 108 views.

Traffic to the BuzzFeed quiz.

We’ve also received some support from Hubway itself. Last week Kendra tweeted out the link copying both BuzzFeed and Hubway on the link:

Kendra’s tweet, tagging BuzzFeed and Hubway.

Hubway retweeted it:

Twitter notifications about Hubway’s activity.

It garnered:

METRIC	RESULT
Impressions	662
Total engagements	22
Link clicks	15
Likes	2
Detail expands	2
Profile clicks	2
Retweets	2

To produce this data we used Hubways 2014 bikeshare data. In addition, we used a wide variety of additional data sets for comparison points. For example, to compare the cost of commuting in Boston by Hubway, T, and car, we used the current cost of a full-priced monthly T pass provided by the MBTA. For car cost we used 2011 AAA data which put the cost for a medium sedan at 57.3 cents a mile, and calculated out the cost of that car commuting the average 7.6 miles per day for 50 weeks a year (assuming 2 weeks’ vacation). The carbon emissions data was based on. Calorie data was based on a bicycle calculator from bicycle.com and USDA calorie data.

We think this was an appropriate way of approaching the issue – the stereotype of the daily commute is one of a painful slog. We, instead, wanted to portray bicycle riding and Hubway use as the opposite of that – fun, whimsical, AND the smart choice. Testers within the target demographic responded that they laughed, and they learned something.

In addition, we felt like it was the right tone both for the audience we were seeking to target and for the data itself. Hubway data, unlike say the refugee data, isn’t particularly serious or fraught. And, on balance the story of Hubway so far is a positive story. People are embracing the system – we just want more people to embrace that system because doing so will make the system bigger and better as well as more enduring. Years ago, Kendra interviewed a woman working with the National Park system and she said that the fact that fewer young people are going to the parks and staying for less time when they visit is a problem. Without a connection to the park system they’re not going to vote for dollars to maintain it, or for politicians on the basis of how they feel about the national parks. The same is true for Hubway – we need young people adopting the system for its long term longevity. This isn’t to say that we don’t love and want to maintain our older riders – and the quiz does nothing to denigrate that audience.

Map of US highlighting states where bee-dependent food crops are grown.

Save the Bees used an interactive map highlighting bee-dependent food crops and risks over time.

By Almaha Almalki, Sean Soni, Jingxian Zhang, and anonymous student.

Save the Bees presentation slides (PDF - 3.2MB)

Overview

We created an interactive US state map of foods that depend on bees. As the user moves a slider, the year changes, and the map changes color to represent the amount of bee colonies left if bee Colony Collapse Disorder (CCD) continues at its current rate. Additionally, we display how food prices might change, given the amount of bee colonies remaining in a particular year.

We paired our interactive map with an informative display about CCD, as well as pre- and post-interaction surveys. We then went to the Copley Farmer’s Market in Boston and tested our interactive with 11 subjects who were shopping at the farmer’s market. Upon completion of the demo, each participant received a free packet of bee-friendly wildflower seeds, and was encouraged to donate or sign a petition to help save the bees.

Methodology

We got our data from the USDA National Agricultural Statistics Service (NASS), which publicizes data about bee colony numbers in the United States. This data provided bee colony numbers by state over the last three years, as well as the estimated amount of money spent on bee pollination by state (the less wild bees, the more farmers must spend on pollination services). We found that bee colony were rapidly and alarmingly decreasing. In order to estimate the rate of decline of bee populations in each state, we calculated the average rate of decline in bee population for each state, and then used this average percent decrease to extrapolate over the entire period of our demonstration. It should be noted that this is a rough estimate, and obviously other factors will influence bee population, and decline is unlikely to be by the same percentage amount every year. Thus, we were careful to tell our participants that this data was calculated as if bee populations were to keep declining at the current rate.

While the colony decline calculation was relatively straightforward, we had more difficulty calculating the increase in food costs. After much research on projected costs, we found no academic work that contained the data we were seeking. Thus, we decided to extrapolate a prediction based on the dollar amount of bee pollination services used. We found that the amount of money farmers spent on bee pollination services was increasing every year, and we thus calculated the average rate of increase for each food crop, and used these rates to predict the price increases. This is a rough prediction at best, since other factors will play into food costs, and alternative pollination schemes are likely to emerge when food prices become high enough to make them economically viable. Thus, we were careful to explain to our participants that these were projected prices, and the real prices could vary widely in the future.

In order to build our demo, we calculated all of these rates of increase, placed this data into a spreadsheet, and then imported it into our Javascript application. Aside from our simple calculations described above, there was no data cleaning to be done, as the data provided by the USDA was already in a very useable format. By making simple calculations, we were able to turn this historical data into a story about the future. By allowing participants to choose which fruits they personally enjoyed, and only have those appear on the map, we turned a large amount of impersonal data into a story about the participant, allowing them to become more engaged and relate to our story on a personal level.

Impact

While we were brainstorming about ideas for this project, we knew we wanted to tell a story using the bee data, and we knew we wanted to create a map. We spent a long time considering different ways to tell this story, but ultimately decided that we wanted to target people shopping at farmer’s markets, since they were likely already predisposed to care about these issues. With this target audience in mind, we decided to focus on the cost of produce, as this would be a very tangible thing to people who are in the process of spending money on produce. With this audience in mind, we began to think about our goals. Our ultimate goal was to help end CCD, and we came up with three concrete ways to make a contribution. First, we would ask participants to sign a petition to ban neonicotinoid pesticides, second we would ask them to donate to Save the Bees, and finally we would give them bee-friendly wildflower seeds to plant. We also hoped to have them think about CCD over the long-term, and share this information with their friends.

In order to gauge the effectiveness of our visualization, we implemented a pre-demo and post-demo survey on the iPad, and also asked a series of verbal questions. The iPad survey asked the participants to rank three issues (CCD, climate change, and urbanization) in order of least threatening to most threatening to our food supply. The verbal questions at the end of the demonstration asked how likely the participants were to buy organic, plant the seeds, and tell their friends about CCD.

The iPad survey was not as successful as we would have liked, with the majority of participants not changing their answers and consistently ranking CCD second. Some participants ranked CCD as the most threatening issue on the pre-demo survey, most likely in an attempt to placate us (one person told us as much). Thus, when they did the same on the post-demo survey, it was difficult to gauge if their perception had changed. Overall, we found that most people had strong pre-existing beliefs about these issues, and our audience was in general well-educated and knowledgeable about these issues (3 of our 11 participants happened to be MIT graduates). We found that the mention of climate change in the survey tended to derail the conversation, as many participants had strong views about climate change. In the future, we would alter this portion of the survey.

The verbal interview questions were much more effective, and here we got our greatest source of feedback. The first person we spoke with was, by pure coincidence, an amateur beekeeper, and his feedback was especially valuable. He loved our idea, but said he wanted to see more information on CCD available to our audience, beyond the facts we presented. Indeed, this desire was expressed by others, so in future iterations we would bring along informative brochures we could hand out. Other participants noted that they already shopped organic, some exclusively. Almost all of them said they would plant the seeds, and most seemed excited about it. When asked if they would share information about CCD with friends, most took it as a suggestion and nonchalantly acquiesced, as if we were making a request rather than an inquiry. Overall, we received very positive feedback, and with a few tweaks, we believe this could be a viable project on a large scale that could make a significant difference in helping combat CCD.

Browse Course Material

Course Info

Instructor

Departments

As Taught In

Level

Topics

Learning Resource Types

Data Storytelling Studio: Climate Change

Assignments

Materials

Methodology

Goals & Audience

Project Assessment

Methodology

Impact

Q&A

Before questions:

After questions:

Testing Responses

Materials

Writeup

Sample Final Project Ideas

Writeup

Overview

Methodology

Impact

Course Info

Instructor

Departments

As Taught In

Level

Topics

Learning Resource Types