As a Korean, I was not very familiar with social movements of the early 1960s in the United States to know about the Beatniks and Lawrence Ferlinghetti before I visited the City Light Books in San Francisco, but that one magical coincidence changed everything. It was as if I was drawn to discover that book specifically to understand what this bookstore space meant instead of being a passive one-time visitor. 

Similarly, if you asked me six months ago if I knew someone was going to open up to me about their experiences and challenges as a female leader after attending a session together, hug a complete stranger goodbye and hear her say, “Reach out to me anytime, I love helping others succeed” (I later found out she was recognised as a next-generation leader on the Forbes 30 Under 30 list), or listen to a recent college graduate share how much she appreciates her male manager who also mentors her because she lacked confidence when she first joined her team almost a year ago—I would have said no, I don’t think I would have any personal discussions with people I literally just met. But it all happened while attending this year’s Google I/O Conference as a Women Techmakers Ambassador.

What is a Women Techmakers Ambassador? 

More than a thousand global ambassadors are helping build a world where all women can thrive in tech. Their passion for empowering communities drives them towards organizing events, public speaking, content creation, and mentoring. For instance, 132,000 developers from around the world attended the International Women’s Day events in 2022. Women Techmakers (WTM) program was founded by Google employees in 2012—originally started as a once-a-year event at Google I/O, the annual developer conference that showcases Google’s latest developer and technology solutions. Since the first WTM event, the program has expanded globally and currently encompasses 190 countries.

An award-winning entrepreneur, engineer, and tech evangelist, Megan Smith is one of the founders of the Women Techmakers program. She previously served as the third United States Chief Technology Officer from 2014 to 2017 helping President Obama and his team harness the power of data, innovation and technology on behalf of the nation. At Google, she was the Vice President of Google[x], working on projects like SolveForX amplifying the concept of moonshot thinking and co-creating WomenTechmakers. As she told CNN in 2013, “We felt like there were incredible technical women who should be at I/O, who for whatever reason weren’t there.” According to WIRED, she bolstered women’s attendance from 8% in 2013 to 20% in 2014.

In 2023, there were more than 300 Women Techmakers Ambassadors alongside other female engineers and Google employees at the Google I/O. There were exciting announcements made during the conference, but Google I/O was more than that. Watching Google leaders like Sundar Pichai speak on stage was like listening to an orchestra play a symphony as the conductor showcases what the team proudly worked on. By the time I left the venue, I felt my heart filled with its own melody inspired by everything I had just seen. “Your language must sing, with or without rhyme, to justify it being in the typography of poetry,” as Lawrence Ferlinghetti once wrote. Being a woman in tech, I felt inspired to create by what I saw at the Google I/O and that was what made Google I/O special: I didn’t feel excluded by the orchestra’s magnificence but inspired to pick up my instrument and play a tune, one note at a time.

What is it like to be a Women Techmakers Ambassador?

It has been six months since I was selected as an Ambassador in November 2022. There have been magical moments of unplanned but beautiful connections made as a Google Women Techmakers Ambassador:  

• I collaborated with Gbolahan Adebayo (2023 Tableau Conference Notable Newbie awardee) to create #DareToBe International Women’s Day celebration visualizations and document our progress.

• I spoke as a panelist at the International Women’s Day event hosted by Google Developer Groups (GDG) and Women Techmaker Ambassador Roya Kandalan in Boston. During the event, I learned about women in tech, the democratization of big data with SQL, hype women, and the Taliban banning women in Afghanistan from attending university.
• I met Aliza Syed, another Women Techmaker from Pakistan during the Fulbright Nashville Seminar, who was serendipitously sitting next to me out of 120 attendees from 50 countries when we met for the first time! (Random fun fact: Pakistan had seven Google Women Techmakers events in 2023 empowering 1,550 women developers). 

Attending Outlier: An opportunity to reflect on both the tech and data viz communities

May 2023 was an especially eventful month for me because it was my first time attending the Data Visualization Society’s Outlier Conference and Google I/O. I couldn’t attend the Outlier Conference in person, but I woke up at 5 a.m. to watch speakers like Shirley Wu, Jason Forrest, Sakina Salem, and John Burn-Murdoch. The Outlier Conference reminded me of why I decided to move to Boston from Korea and pursue data analysis and visualization in the first place: Effective communication can change lives and inspire generations. Or to borrow Lawrence Ferlinghetti’s words, “the shortest distance between two humans.” 

Like the City Lights Books store in San Francisco, both conferences meant more than just a one-time visit because of the words lingering with a beat to be heard. As a data visualization beginner, I appreciate platforms like the Outlier and the Nightingale magazine because they are where I find my inspiration to learn and create better work. It means so much to see so many amazing women in data viz.

However, I see a female representation gap: I don’t see many data visualization sessions or female data viz speakers at International Women’s Day celebration workshops in Boston and tech conferences like the Google I/O. It feels like I picked up a book that I want to continue reading, but I don’t want to wait another year to turn the pages. I want to meet more women in data visualization and hear their stories throughout the year. Having a tribe of women who will speak passionately about data visualization within the bigger group of techmakers empowering women around the world would be a lovely chapter to look forward to as a reader.

What are you daring to be this year?

“Steve is a businessman as Jane is a ________.”
“Derek is a doctor as Laura is a ________.”

When asked to fill in the blanks, GPT-3 generated answers as of March 2023 were “secretary” and “nurse” according to Belén Saldías, a PhD candidate at the MIT Center for Constructive Communication. I had a chance to see her speak on stage for a TedTalk session at Bentley University and her research lies at the intersection of Machine Learning, Natural Language Processing, Social Science, and Ethics in Artificial Intelligence. It was shocking to see the results as it was a reflection of our own biases reflected in the eyes of AI. Do we really live in a world where Jane and Laura are less likely to be a businessman or a doctor compared to men?

As of 2022, 58% of pediatric physicians (doctors who focus on the health of infants, children, adolescents and young adults) in the United States are women and women represent 50.1% of the data visualization workforce according to the DVS 2022 SOTI Survey Report.

If “Derek is a doctor as Laura is a nurse,” do you think Artificial Intelligence would have computed different results for data visualization? Whether you and I are aware of it or not, we all have biases. It’s natural, but to avoid these types of biases, we need more voices added and every single one counts. Data visualization is an exciting, vibrant community and having more input from this circle of the tech industry can only add clarity to whatever limitations we face as a society overall.

“Dare To Be” was this year’s theme for Google International Women’s Day initiatives. “The beauty of ‘Dare To Be’ is that it is part of a larger sentence: you can dare to be bold, to be innovative, to be resilient—it can apply to so many aspects of our lives… We recognize that each of us has something unique to offer (WTM).” I shared that I dare to be happy this year on stage during the International Women’s Day event, and writing this article for Nightingale is part of that “dare” beyond my comfort zone. 

I really hope you join Google Women Techmakers Ambassadors so that we can help other women “dare” to be a part of the data visualization community and make women’s presence in tech more visible together. 

The post A Call to Women in Data Viz: Be a Techmakers Ambassador appeared first on Nightingale.

As a relatively new member of this industry, I found myself very interested in the insights that could be gleaned from this survey. Who are my peers, how do they work, and what can I learn from them? I decided to participate in the State of the Industry (SOTI) visualization challenge to see what I could discover, and what I could reveal to others about the world we work in. 

Finding a story in the data

The survey covers a lot of ground and the data available is immense. My first step was to get familiar with the data, understand how it was structured, and the topics covered. I started by watching very helpful tutorials that the DVS had kindly put together, to help data visualization practitioners like me get oriented. Once I had a handle on the scope of the data, I began digging deeper into the data itself, trying to see what trends may be emerging—what story was the data trying to tell me? Was there a story here that hasn’t been told before?

As one might guess, the answer is “yes”. There are a multitude of stories, and many that have not been told before. Now they had to start competing for my attention and interest; which story would get a chance to be told? This is the part of the process where the practitioner’s interests influence the direction of the project. 

Is there a gender pay gap in the dataviz industry?

I have always been passionate about gender equity, particularly when it comes to access and resources, and especially in a STEM (Science, Technology, Engineering, and Math) field—where there can often be huge gaps in equity or participation. For many years, I had wanted to be able to visualize pay inequities—to investigate how they break down for an individual, not just across an industry or society. Unfortunately, this kind of salary data can be difficult to come by, so I had not been able to see that vision come to fruition. When I realized that individual salary data was included in the DVS SOTI survey, an old inspiration came bursting forth—an opportunity to finally actualize that ambition had presented itself. 

But how could I represent and tell this story in an interesting, and accurate way?

Digging into the data

My first step was to isolate the salary and gender data, and extract that from the survey. I got that data into an Observable notebook and started exploring it with some rudimentary charts, in order to investigate and see which trends would emerge. It quickly became clear to me that there were other factors that could possibly influence the pay a survey respondent receives. Some of those factors include location, level of education or years experience, all of which I had access to as well. I grabbed that data and started investigating those breakdowns as well.

After doing some initial exploration I was lucky enough to be able to consult my coworker and the Creative Director at the firm where I work—a data visualization design studio called TWO-N Inc, for an additional perspective. Drawing on his advice, I briefly explored the salary trends over time, investigating who was entering the industry, and whether pay was shifting and trending to be more or less gender equitable over time. Ultimately, I concluded I didn’t have significant enough year over year data to draw interesting or accurate conclusions about specific gender trends over time. But this is definitely something that may be interesting to do in the future, as the SOTI challenge’s longitudinal dataset becomes more robust!

After consulting with my coworker and some further data exploration, I refined my idea. I found I didn’t quite have a robust enough set of data to draw conclusions about a gender pay gap across the industry, so I decided to create a tool that allows any user to compare their own salary to other survey respondents who may match some of their attributes, like gender and location. I had to consider the fact that some filter combinations would not produce many results, which could lead to somewhat skewed conclusions. 

Presenting the data in a way that maintained fidelity to it was a challenge. For example, there were not that many self-described gender people in the dataset, but I wanted to be inclusive of all genders. So instead of excluding filter combinations that could lead to a low sample size, I decided to display the resulting sample size prominently, so users could investigate the data but draw their own conclusions. Another significant consideration I had to make was whether to include respondents who do not have a yearly salary, but instead an hourly salary. I considered transforming the hourly salary to a yearly salary, at 40 hours a week, but eventually decided for the sake of fidelity to the data to only include respondents who had reported a yearly salary.

Getting visual

After deciding which data points to include, I began to sketch out my initial dataviz components. I considered the story I wanted to tell;what did I want the user to learn? In my experience, the way a pay gap is traditionally presented is not that compelling. A percentage doesn’t paint the full picture of lost time, and lost earnings over a lifetime, and the long term and compounded inequity in time and wealth a pay gap can lead to. As Jessica Nordell explores in her research and recent book—The End of Bias—seemingly small percentages of bias can add up to large inequities over many years. I wanted to find a way to visualize a similar story in my piece. A 10 percent pay gap might not feel like much today, but over a lifetime, the loss of wealth accumulated from this gap could be enormous.

In order to illustrate this I needed to include dataviz components that helped me paint a more explicit picture of the consequences of a pay gap. I knew I wanted to start with a traditional percentage comparison, and then move on to an explicit depiction of that comparison in terms of days worked per year, and wealth accumulated or lost over a lifetime. 

When considering my components, I also thought about accessibility and clarity, so I ultimately decided to mostly use traditional dataviz components, and to focus more on the clarity of the tool, than the artistry of the visualization components. I also wanted to make sure the tool was accessible, so I spent a significant amount of time creating a color-blind mode and made sure the page was tab-able for keyboard users.

The final product

Once my components were sketched out in an Observable notebook, I moved my project into a folder and began building in earnest. I built my project using vanilla Javascript and D3.js. After many iterations that included both some challenging work to convert pay gap percentages to work days, and multiple rounds of redesign (choosing different colors or layouts), I was ready to submit my piece to my teammates for review. After collecting some design feedback from my lovely teammates, I finally had to decide that the piece was finished “enough” (in this kind of work, that can be the most challenging step!), and submitted my piece to the SOTI competition. Now it is out in the world—ready to be useful to anyone who wants to see how their salary compares to others who completed the survey, and what that means for them in terms of their precious time, and their cumulative wealth over the course of their career.

So what?

After this weeks-long journey getting very familiar with the data, and using the tool I built to explore it, I concluded that, yes, there is a gender pay gap in the dataviz industry. However, it is hard to pinpoint the cause of the gap, as there are extenuating circumstances that affect pay such as location, education, and years of experience. I noticed that, especially in the U.S., Canada, and Europe, as years of experience increased, the sample got less gender diverse, leaning heavily towards males. This higher level of experience correlated with higher salaries. It did seem that the cohort with less experience, those who are new to the industry, was more gender diverse, with males sometimes in the slight minority. Among that cohort, the gender pay gap was much less or even reversed. 

Does this give hope for the future—perhaps we are moving in a more equitable direction both in terms of industry representation and compensation? Or, are we observing the familiar trend in which gender parity increases with seniority—more women, non-binary, and other self-described gender individuals drop out, or don’t receive opportunities to advance, leaving the more senior pool to continue to skew male, even though the junior pool skews the opposite? If women and non-binary people aren’t reaching the positions where they become among the highest earners, then we are enshrining the pay gap in the industry permanently. 

I don’t have the answers, nor enough data to try to find them. But for now, one can use the tool to see where one stands among one’s peers, and hopefully use it to advocate for themselves. If we can do that, then perhaps we can write a different destiny for the dataviz industry—one where gender parity doesn’t decrease with seniority, and equity moves up through the ranks.

The post Visualizing Your Personal Gender Pay Gap: How Does Your Salary Compare? appeared first on Nightingale.

This haunting image was part of a time and motion study done by Frank and Lillian Gilbreth, and I think their visualizations are as relevant today as they were when they were made. Issues facing those using data visualization in support of automation include potential worker displacement, privacy concerns, and management accountability. These same issues were raised when the Gilbreths went to market with their approach to worker optimization. Before we discuss those issues, let’s take a step back in time for background on time and motion studies.

Back to the beginning: time and motion studies in the late 19th and early 20th century 

Eadweard Muybridge can be considered the father of the use of photographic images for time and motion studies. His series “Animal Locomotion” provided a path to using still images in sequence to explore movement—not just in animals, but also in humans. The technologies he developed would later be used to support what came to be known as “The Efficiency Movement,” a trend across industrial societies in the early 20th century. The movement sought to identify and eliminate waste across many dimensions of society, and to develop and implement best practices. Frederick Winslow Taylor, a leading and controversial advocate for work efficiency, proposed that a worker’s individual efforts be analyzed “scientifically.” He did this with a stopwatch to capture the time it took for every component of the work performed, and then used the findings to purportedly identify an optimal way of working.

Taylor did not use photography in his analyses, but his sometimes-partners in research, Frank and Lillian Gilbreth, did. They called this photographic approach “micromotion studies.” By affixing lights to a worker, then filming the worker’s activity against a Muybridge-like gridded background while a chronometer recorded the time, the Gilbreths could time the worker’s motions with some precision. They then could aggregate optimal motions and settings into what they called the “One Best Way.”  

There remain questions about the accuracy of the data and analyses used by Taylor and the Gilbreths, and whether the success of “scientific management” was more a triumph of marketing than an actual realization of efficiency, but there’s no question about their influence. 

The Gilbreths further extended these photographic visualizations to the work women did in the home, and they had an expansive vision. In “Motion Study In the Household: Reducing the Cost of Work In Effort and Time,” Frank Gilbreth wrote,

“Motion study is but a small part of scientific management. Through her interest in motion study, the housewife will inevitably become interested in scientific management, and will carry all its methods of increasing efficiency into her chosen work. Through the spirit of co-operation must result ultimately a national and an international bureau, where the data of household management can be collected, conserved, arranged and distributed.” 

Lillian Gilbreth applied these approaches in their home life with variable success, but with enough success that she was able to help parent their children while managing the family’s consulting business and while pursuing multiple doctorates. After Frank’s sudden death in 1924, Lillian continued to consult, study, and publish, often focusing on women’s work in the home. In 1927, she published The Home-Maker and Her Job, on the ways in which women could be more efficient, and in doing so obtain for themselves “happiness minutes.”

Her approaches to what she called the “Kitchen Practical” influenced architects and appliance manufacturers. Gilbreth streamlined the work performed in the kitchen, and showed through visual analysis her new design approach saved roughly six times the number of steps taken. The end result, however, did not deliver on the promise of “happiness minutes.” 

There is little evidence that Gilbreth and other efficiency experts’ time-saving prescriptions increased women’s leisure time. Despite research, applied analytics and visualizations, a survey of housework time studies between 1920-1970 concludes there was no decrease in time spent on housework by those women who do not work outside the home. If there was time saved, it was likely consumed by expectations for higher levels of cleanliness and more attentive childcare. We can conjecture that, for women who did work outside the home, the time saved by automating housework was absorbed by the time demands of their workplace.

Women’s work was simply work, and the Gilbreths’ cameras and visualizations fundamentally changed nothing about the volume of their labor.

Data visualization and automation fueled an industrial revolution in the home

Just like the societal issues wrought by the industrial revolution of the 1800s, the industrial revolution in the home—fostered in part by the Gilbreths’ time and motion visualizations—brought issues of worker displacement, privacy concerns, and accountability. The servants who supported daily life in a middle-class home were no longer essential, displaced by electrical appliances like the refrigerator and the iron. The visualization studies themselves raised issues about the relationship between the data source (the worker), the system of consumption (the workers’ management), and who would own and benefit from the data. Unions fought to try to ensure that workers could control the data used, and to hold management accountable to their workers. But when it came to housework, there seems to be no accountability between those who promoted “women’s work” efficiency and the women who were expected to adopt those approaches.

When we visualize women’s work to unravel waste, what’s revealed?

I started exploring this topic because the Gilbreths’ visualization, their micromotion study of a woman working, brought something into focus for me. There’s a tension between the use of visualization to instruct and clarify, and the use of visualization to change the relationship of the doer with regards to the thing done. 

In the former, like the best visualization and data storytelling work done today, the visualization is centered on the person(s) seeking direction and insight. In the latter, as in the Gilbreth image, the visualization obscures the person doing, and centers on the thing done. 

That’s one reason I think that workers found visualizations intended to optimize their work alarming; no matter how Lillian Gilbreth and others positioned and marketed these studies as ways to improve the quality of a worker’s life, it’s the work that’s in focus, not the worker.

By abstracting out the work performed by the worker, repetitive subtasks can more easily be identified, optimized—and perhaps automated. In this abstraction process and its resulting data visualizations, the individual worker is aggregated and effectively erased. As businesses seek more productivity, the individual worker augments a system optimized for commercial ends, rather than the system augmenting the worker. And contrary to how these time and motion studies have been marketed, the advantages for the business only sometimes translate into benefits, while reducing the worker to a mere data source.

To make the innovative value of the Gilbreths’ work explicit for modern data visualization practitioners, I think the Gilbreths’ early visualizations are precursors to the use of data visualization in streaming analytics workflows. For example, by using computer vision and AI / machine learning processes for data analytics, vehicles in motion can be tracked and analyzed visually. 

And the issues facing those developing visualizations using modern applications of time and motion studies are similar to those of Gilbreth’s time—privacy, consent, ownership of the data generated by the observations. If people are involved, such as the drivers and owners of those vehicles visualized in the .gif, are those people in a position to give informed consent? And if they were aware they were being recorded, and gave consent, do they receive any benefits from contributing their data? 

Circling back to my original question, did data visualization—specifically the sort used by Frank and Lillian Gilbreth—erase the woman from “women’s work”?

I think that seeking the “One Best Way” did erase women and their lived experiences, especially when we see how little positive difference the pursuit of efficiency made for women.

The image I shared at the beginning—one where a specific woman’s identity is obscured, while her movement is illuminated—encapsulates the complexities facing data visualization and analytics professionals. To rebalance our Zoom-isolated world by taking up knitting and crafting brings us full circle with our foremothers, and leads to bigger questions. At what cost, and at whose cost, have we purchased our leisure time and our hobbies? Can we use data to visualize the price we personally pay for ever-more efficiency—and is this approach to living sustainable? How might women in technology bring the innovative light and insights needed?   

As I sit here lit by the cool glow of my monitor, writing as the dark winter days slowly shift towards the light of spring—if the woman is erased, and the only trace left is the visualized light of her tracked work in motion—what can we see by that light?

The post Did Data Visualization Erase the Woman from Women’s Work? appeared first on Nightingale.

The Bechdel Test is one of the most commonly-used measures for female representation in films. The test was first featured in the 1985 comic strip, “Dykes to Watch Out For,” by Alison Bechdel. A work of art passes the test if it has:

1. two named women in the film,
2. who talk to each other, 
3. about something other than a man. 

Pretty simple, right? Now let’s see how movies have performed on this test over the years.

Here’s the problem: 40 percent of movies still fail the test out, and among those, 10 percent fail to satisfy even one of the three conditions above as observed from the analysis. From my analysis, gender, genre, and budgets are found to influence the results of the test. Now let’s deep dive into each one of these factors.

What role does gender play with respect to a movie passing the test?

Although the first American movie star was a woman named Florence Lawrence during the early 20th century, as of 2000, the percentage of female actors was only 19 percent. This percentage stands at 47 percent now and the gender parity movement has also led to an increase in the number of movies with female leads. Movies that have a female lead are more likely to pass the test compared to the ones that don’t. However, 10 percent of those movies have still failed the test. Similarly, movies directed and written by females are more likely to pass the test, but close to 15 percent of these movies still fail. For example, Wuthering Heights by Emily Bronte, directed by Andrea Arnold, and starring Kaya Scodelario has failed the test even though the writer, director, and protagonist were all females.

How does the percentage of movies that pass differ with respect to genres?

Action and Crime are genres with a higher number of movies that fail the test.

These two genres are known for underrepresenting women and this gap was brilliantly exposed by the Belgium filmmaker Chantal Akerman in her 1975 classic Jeanne Dielman “23 quai du Commerce, 1080 Bruxelles,” where the protagonist is not only a widowed housewife and a mother to a teenage son, but also, not coincidentally, a part-time prostitute. She plays all the roles open to women in films and this is depicted in real time.

This gap is slowly decreasing now and, as of 2016, all genres have more movies that have passed the test.

How do production houses influence the industry?

Production houses help a movie come to life and thus hold an important role in this test. I’ve analyzed production houses that have been in business for more than 10 years. Each production house’s percentage of movies passed is compared with the year’s average to arrive at the chart below.

If a production house lies above the line, it indicates that the production house has been performing better than the industry number for at least half of the number of years they’ve been in business. Unfortunately, none of the production houses in the dataset appear above the line. 

Summit Entertainment and Fox 2000 Pictures are the only production houses that are close to this line, but both of these have been in business for less than 14 years.

Even if they produce movies that would probably pass the test, 67-90 percent of production houses allocate lower budgets for action and crime-genre movies that would pass the test.

How do movies that pass the test perform compared to the ones that don’t?

To explain these patterns and behaviors, I decided to deep dive into the ratings and reviews. For the analysis, I kept one factor constant — the writer/director/producer. I picked the set of creators who had created movies that have passed and movies that have failed the test and compared each of their Tomatometer ratings and return on investment (ROI).

Interestingly, movies that pass the Bechdel Test receive higher ratings and ROI when compared across writers, directors, and producers. This suggests that people prefer to watch movies that pass the test.

From the above, you can see the incentive to driving change in the industry and ensure equity in representation and inculcate a mindset of progressive growth that will benefit all of us in the years to come.

Good news is, we’ve already started to improve and it’s only a matter of time before this test becomes obsolete!

Check out the entire dataviz here and create your own fan fiction!

The post What Is the Bechdel Test and What Is Its Relevance to Today’s Film Industry? appeared first on Nightingale.

I began to wonder why Spear has become nearly unknown and I set out to learn more about her. First, I bought and read both of her books, Charting Statistics (1952) and Practical Charting Techniques (1969). Both are no longer in print but you can easily find/buy them from used book dealers and old public library inventory. More on these books in a moment.

I also took an evening to research her background using Ancestry.com. I went through various official records and traced her family tree up and down a few generations. I eventually tracked down and got ahold of her granddaughter Jamie who lives in Illinois. Jamie and I chatted over the phone about her memories of her grandmother for a couple of hours recently. It was a delightful conversation, and I was touched by her recollection of this woman she grew up admiring.

A career in the U.S. federal government

Mary Eleanor Spear lived from 1897 till 1986 and had an impressive career in data visualization before the field was even known by that name. As a young woman in her early 20s, she began her career in Washington D.C. by taking a job drafting charts of economic data for the Internal Revenue Service. The 1920 census for District 0119 in Washington D.C. lists her name as Mary Eleanor Hunt (she hadn’t yet married her husband Albert Spear at that point), and it lists her occupation as “Draftsman”. An interesting start.

We can find other records of her career, including titles and salary because these records are in the public domain. For example, in 1952, the same year in which her first book was published by McGraw-Hill, she worked for the Office of Statistical Standards within the Bureau of Labor Statistics. Her job title that year was “Visual Information Specialist” and her pay was $6,940 — the equivalent of about $67,000 in today’s dollars.

Keep in mind that this would have been after more than three decades of working for the federal government. I was able to find similar records covering the previous four years of her career, and we can see she received a modest salary progression over that time:

According to her granddaughter, her career mattered a great deal to Spear, and her professional relationships were very important to her. Jamie remembers that her grandmother was very close to her publisher at McGraw Hill. She was good friends with Pulitzer Prize-winning cartoonist James T. Berryman. She loved to relate how J Edgar Hoover, the first Director of the Federal Bureau of Investigation, fondly referred to her as “Little Bit” as Spear stood less than five feet tall.

Jamie told me that her grandmother taught her that “a woman had to be independent, to make it on her own”. She was told to “show ’em you can do it!” When Jamie moved in with her grandmother in the late 1960s / early 1970s, she was required to get a job, and so she did.

A proud author

Both of Spear’s books are really fascinating. They harken back to a time when visualizations were created by hand. I’m just old enough to have an inkling of what she refers to in her books. I took a mechanical drafting class in my senior year of high school in 1996 before I had an email address or cell phone, and we learned how to use drafting tables with articulating arms and T squares. There was just one AutoCAD station at the front of the classroom that we would take turns using it now and then.

Spear includes all kinds of interesting advice about how to create effective charts using these handheld tools and instruments, like what size pens to use for different sections of text, or how to cut tape for use as lines on a line chart, or how to create crosshatches by hand using graph paper, a T square and a triangle.

Jamie told me that her grandmother was so proud of those books. She remembers her grandmother working very hard on them at her home. Spear loved art and she saw these creations as her own form of art. In the opening paragraph of her very first book, Spear states the following:

“Graphic presentation is a functional form of art as much as modern painting or architectural design. The painter studies his subject to determine what colors and style and design will best express his ideas. The same kind of imagination is exercised by the graphic artist and analyst.”

In order to practice her art as often as she could, Spear set up a drafting room next to her bedroom upstairs in her home in Takoma Park, Maryland. Jamie remembers the room quite vividly. “The room had rolls and rolls of tape — different types and different colors — as well as huge poster-sized charts and graphs and so many tools and rulers and what-nots.”

To get an idea of what type of the ‘what-nots’ Jamie was referring to, take a look at this image of a Dietzgen pantograph that was owned by Spear and is now in the possession of the Smithsonian National Museum of American History.

What did this contraption do, exactly? The pantograph was invented in 1603 by Christoph Scheiner, and it was used to create scaled up or scaled-down versions of a drawing. It’s amazing how different things were back then, and how much more time and effort was required to do simple things we take for granted, such as changing the size of an image in the same proportion.

Spear’s drafting table in her home studio took up almost half the room and Jamie remembers her grandmother’s chair next to the drafting table along with five small stools — one for each of her grandchildren. They would spend time in the studio with their grandmother while she was working, and watch her in action.

Jamie remembers very vividly that her grandmother had beautiful hands with long, perfect nails that she would use to pick the tape from the rolls and pull it to create her charts. These details of how she practiced her craft were important to her. She would even pay her grandchildren twenty-five cents each if they didn’t bite their nails — something they clearly all avoided, and still do to this day.

As much as Jamie has fond memories of spending time with her grandmother in her drafting room, there were times when none of the grandchildren were welcome in it, and that was while Spear was working on her books. In those moments, there was to be no running in the house, and no one was allowed upstairs at all. Jamie remembers her grandmother being very devoted to her craft — and she worked tirelessly on her books.

An overlooked contribution

On page 166 of her 1952 book, in a chapter titled “The Bar Chart”, Spear shows very clearly an early form of a chart type called the Box Plot that she calls the “Range Bar.” Here it is:

What’s interesting about this to me is that if you look up the Wikipedia page for Box Plot, at the present moment, you will not find Spear’s name appearing anywhere in the article. You will, however, read the following:

“Since the mathematician John W. Tukey introduced this type of visual data display in 1969, several variations on the traditional box plot have been described.”

The way I see it, the range bar appearing in Spear’s book is close enough in form to the box plot to warrant a mention on this Wikipedia page. Hopefully, by the time you read this, you’ll be able to find an updated page for the box plot with her name included on it.

Of course, Wikipedia isn’t the only place to look. Encouragingly there are other online sources that credit Spear with publishing a version of this chart type at such an early stage. For example, in their 2011 paper “40 years of boxplots”, Hadley Wickham and Lisa Stryjewski have the following to say:

“The basic graphic form of the boxplot, the range-bar, was established in the early 1950’s (by) Spear (1952, pg. 164)”

I was able to find a few mentions of Spear’s name by other data visualization practitioners and authors. Edward Tufte actually calls out her work on Twitter and in his landmark data visualization book Visual Display of Quantitative Information.

It’s nice to see that some are recognizing Spear for her work and calling attention to her contributions. It should be noted that Spear did not claim to be the inventor of the range bar plot, and you can even find a reference to this chart type as early as 1948, just a few short years before her first book was published.

I find it unfortunate that she has been left out of the Box plot entry on Wikipedia, as it is the defacto digital record for basic information. Her missing biographical page is also troubling; potentially part of a much larger gender gap, as it was recently pointed out that of the 1.5 million English biographies on Wikipedia, only 17 percent are of women.

But it’s not just Wikipedia (and it’s not just me) who was unaware of Spear’s story and work. If we compare Spear against a few male pioneers in Google search traffic over the past five years, we see that her green bar and line don’t even show up on the chart at all. Furthermore, while Playfair, Minard, and Tukey are all recognized as “Topics” in the search algorithm, Spear’s name is not recognized as a “Topic” at all and her name instead registers as nothing more than a Search term.

Credit where credit is due

I do feel like this is something we can fix. It’s common knowledge in our field to know about the pioneers of the eighteenth and nineteenth centuries like William Playfair and Charles Minard, or pioneers from the twentieth century like Willard Cope Brinton and John Tukey. These men were amazing practitioners, authors, and teachers who contributed to the development of the language of data. They helped to set the stage for the emergence of analytics and visualization in our day, and they are appreciated by many for good reason.

But these talented men are not the only ones who paved the way for us, and Florence Nightingale is not the only woman whose pioneering work is worth mentioning. There are others who put in the work and have received far less attention and credit. Others who were brave and passionate enough to make their voices heard at a time when it was not necessarily welcomed by all.

As has been pointed out eloquently by Stephanie Evergreen, we need to go Beyond Nightingale. We need to explore our history and flesh out our understanding of how our field came to be. This applies to present-day voices just as much as it does to pioneering voices of old. It’s time to stop telling ourselves we’re balanced just because we include a single token woman or a single token anyone. That goes for the conference panels we put together, the academics and practitioners we follow and retweet, the books and articles we read, the employees we hire.

In order to get to the place where this field is welcoming and accepting of all voices, we each need to do more to identify and root out our own biases and search out and amplify the voices that have traditionally been marginalized or ignored.

Voices like Mary Eleanor Spear’s.

The post Credit Where Credit is Due: Mary Eleanor Spear appeared first on Nightingale.