Over half of women CS students are told their gender gave them an edge
A recruiter heard the number and assumed it was backwards: 55% of students who identify as women say they have been told their gender unfairly helped them win job opportunities. Stephanie Tang, CEO of The Percentage Project, which surveys computer science students across 20-plus campuses, shows how climate data like this exposes gaps in belonging. The question is what anyone can actually do with the numbers once they have them.
See the full survey findings and a three-step inform, reflect, advocate framework for using them
Stephanie Tang, CEO of The Percentage Project, presents with co-presenter Luna the climate survey data the nonprofit collects from computer science students across North American campuses. The data shows gaps in imposter feelings, unfair-advantage comments, microaggressions and classroom comfort by gender, race, gender identity and income. The talk then teaches a three-step model of data-driven advocacy (inform, reflect, advocate) for turning those numbers into change.
“Some might argue that these feel like opposing forces, but I want to make it clear that these are not mutually exclusive.”
“It's important for you as a consumer of data to know where your data is coming from.”
What you will learn
Apply the inform, reflect, advocate sequence to turn survey data into concrete change.
Use a matrix of advocacy steps by audience to brainstorm actions, starting with informing students.
Spot gaps in belonging that summary statistics hide by breaking results down by demographic group.
Design climate surveys that protect participants through confidentiality, voluntary participation and excluding groups under 1%.
Connect statistics to faces and personal stories so awareness spreads by word of mouth.
Key takeaways from the speaker
Embracing Diversity and Feeling a Sense of Belonging
Power of Words
Inform, Reflect and Advocate
Details
Level
All levels
Format
Workshop
Technical depth
Non-technical
For
Student, Early career
Length
37:33
Published
June 2023
Questions this talk answers
What is data-driven advocacy?
The speakers define it as three steps: inform others by presenting data in a digestible way, help them reflect through platforms for critical discussion, and advocate by using the data to make real change.
How does The Percentage Project protect survey participants?
Responses are confidential and voluntary, demographic groups under 1% of the population are excluded from aggregate analysis, data is spam filtered, and only aggregate data is released as a benchmark.
Do imposter feelings differ by gender among computer science students?
Yes. As the speakers cited, 18% of men, 27% of women and 44% of non-binary students often or always feel their success is the result of some kind of error.
How did a campus project grow to many universities?
A community photo gallery pairing student portraits with statistics spread by word of mouth, and students at other schools asked to bring the project to their campuses, leading to an annual coordinated social media campaign.
Can the data be analyzed at intersections of demographic groups?
The team said intersectional analysis is an area it plans to work on in the coming year, with results posted on its website and Instagram.
About this talk
Introduction to The Percentage Project and our Speakers.
Details about Our Approach, Our Data, Methodology, Metrics, Demographics and Survey Results.
Exaplanation of Data-Driven Advocacy for Students, Faculty, and Working Professionals and details on how to get involved.
Diversity and belonging are two sides of one coin (68 s)
The stat a recruiter thought was backwards (51 s)
Inform, reflect, advocate: data-driven advocacy in three steps (90 s)
Speaker 1: Welcome to the Exploring Data-Driven Advocacy session here at the Women Tech Global Conference. We're super excited to have you. And today we'll be talking a lot about how we can explore data-driven advocacy and use data to empower people of diverse backgrounds, as well as departments and changing the culture around underrepresented students. Speaker 1: So to guide us through this, here's our agenda for the day, which can be split broadly across who we are, our data, and our findings, and how we can really translate data into action through data-driven advocacy.
Speaker 1: Starting with who we are, I think it's really important that when we're talking about data, we disclose who we are, what our mission is, and why we're doing this. Speaker 1: It's important for you as a consumer of data to know where your data is coming from. Speaker 1: So without further ado, who are we? To introduce myself, my name is Stephanie, and I'm the CEO of The Percentage Project. I'm also a software engineer at Amazon Robotics.
Speaker 1: And I'm Luna. I'm the chief outreach officer at The Percentage Project, and I'm also a software engineer at Nourish. Speaker 1: So we are part of The Percentage Project, an organization, a 5001c3 nonprofit dedicated to helping every person of every background embrace your diversity and feel a sense of belonging in the field that they pursue.
Speaker 1: I've highlighted two key parts of this message. The first is to embrace your diversity, which is our differences and what makes us special, what makes us unique. And the second, a sense of belonging. Speaker 1: Which is to include everyone to bring people together and to feel that you truly belong. Speaker 1: And some might argue that these feel like opposing forces.
Speaker 1: Some might argue that these feel like opposing forces, but I want to make it clear that these are not mutually exclusive. Speaker 1: Rather, they are two sides of the same coin. We want to bring people together, while also making feel people included, but not through erasure of their identity or pretending that we don't see color or we don't see gender.
Speaker 1: It's through understanding their backgrounds, their life, and who they truly are as a person. And that is what true acceptance and belonging means to us. And that's our guiding mission for today. Speaker 1: So we'll talk a little bit about our approach. Today, we'll be talking about, you know, our approach to data driven advocacy at the percentage project. We can define this approach in three prongs.
Speaker 1: The first is students. We believe that students should have an active say in shaping their educational experience, and so they should be informed. Speaker 1: And how do we inform them? Every year we work with student leaders to collect climate data on their campuses. Speaker 1: So we go beyond, you know, what students are of X demographic. We try to delve deeper and learn about how they actually feel around their peers in their class, their professors and within their community.
Speaker 1: The second prong is departments. It is absolutely critical that the data that we collect helps to drive concrete progress. Speaker 1: And so we believe in keeping departments up to date, detailed analysis of their climates, their campuses, and giving individualized data reports and dashboards to help them identify key areas and challenges within their departments and help fuel data driven decisions that can actually alleviate and Speaker 1: The current circumstances and eventually close the gap.
Speaker 1: Finally, it's community. We believe that we are stronger together, and our focus is on connecting campuses from all over to advocate for equity in computer science. Speaker 1: It's in this vein that we also aggregate all the data from all the different participating campuses that we have every year to release a benchmark. Speaker 1: And this is where we have an aggregate benchmark that each school can use to then compare their own personal and individual school data towards the benchmark and then drive accountability in technology education.
Speaker 1: So to get started, we're going to show you firsthand a little bit about what we do and the questions that we ask and the data that we collect and analyze. Speaker 1: So if you can, and those who are participating in the chat, if you could join slido.com on your phones and just enter this number 8 0 5 6 0 9 2.
Speaker 1: I'm going to give a couple couple moments for everyone to just join and check it out, and then you should be able to interact with a poll. Speaker 1: Yeah, it sounds like we could probably move to the next piece. Speaker 1: So if anyone really still needs it the slido slido numbers still on the side of this of the screen.
Speaker 1: So now you should see on your phone that the poll is open. Speaker 1: Note that this is anonymous so don't worry, you can you can put whatever you want here. Speaker 1: We won't know who who said what just like in our survey. Speaker 1: So the question here is sometimes I feel or believe that my success in life or in my job has been the result of some kind of error.
Speaker 1: So it would be great if you guys could give it a shot and answer where rarely or never. Speaker 1: Sometimes or often or always. Speaker 1: Like we have some response. Speaker 1: A couple coming in for rarely or never and sometime. Speaker 1: So I'm just going to give another moment for anyone who hasn't been able to participate yet just to participate.
Speaker 1: Okay, great, so we can see now. Speaker 1: Right now, I think we have a couple participants, and it looks like about two thirds feel rarely or never, and a third feel sometimes. Speaker 1: Okay, so this is kind of our mini poll or mini survey within this group let's check out what the stats were for across 1500 students from this past survey year.
Speaker 1: Awesome so you can see here that about 53% said rarely or never about 22% said sometimes and about 24% said often or always. Speaker 1: So this is our summary statistic you can see that this is data that's just across all of our survey participants for the past year and and the data isn't broken down any further so let's actually now go and break this down a little bit further.
Speaker 1: So you can see here same graph on the on the left side, the summary statistic that we had now on the right, you can see us actually breaking it down across a specific demographic in this case gender. Speaker 1: So you can see here now we can start to analyze the differences that we're noticing across students on various campuses that have worked with us.
Speaker 1: So you can see here that for those who identified as men 18% said often or always versus 27% for those who identified as women and 44% for those who identified as non binary similar kind of trend or trajectory with. Speaker 1: The sometimes which is starts at 19.76% for those identified as men 25.64% for women and 22% for those identified as non binary.
Speaker 1: And so you can see how this this trend occurs and we can see kind of the gaps that start to form across different demographic groups. Speaker 1: So let's try this again and do another another poll, so if you go back onto your phone and check out. Speaker 1: The question is, I often compare my ability to those around me and think that they may be more intelligent than I am.
Speaker 1: So give give the poll a little shot. Speaker 1: So let's try this again and do another poll. Speaker 1: Okay, great. Speaker 1: So we can see here that right now we're looking at something like 50%, sometimes 50% often or always. Speaker 1: So let's just go and check out the data that we got from the survey.
Speaker 1: So you can see here that often or always is about 60.9%, sometimes it's 22.7% and rarely or never was 16.4%. Speaker 1: And again, let's break this down into a demographic. Speaker 1: In this case, let's break it down again into gender. Speaker 1: And so you can see here that often or always for those who identify as men has Speaker 1: has been 50.59% versus women, which is 69.27% versus non-binary, which is 76%.
Speaker 1: So you can see a gap from the lowest to the highest percentage is around 26 percentage points. Speaker 1: And then the gap between men and women here is around 19%. Speaker 1: So very large gaps here in in feeling whether or not that you're comparing yourself and your own abilities to others and feeling that other people are more intelligent or being worried that other people are more intelligent than you.
Speaker 1: And this is a classic question. Speaker 1: That's part of the actual tests that people do for for first checking if they if they're worried about other people's beliefs or feelings for them. Speaker 1: So this is actually a question that's in a lot of research as well. Speaker 1: So next slide. Speaker 1: Yeah, so now kind of move on to dive a little bit more into our data.
I know you've seen some of it already, but I'll talk more in depth about our methodology and then kind of see more questions that we survey across different demographics. Speaker 1: So first discussing kind of our methodology. Speaker 1: When it comes to data collection methodology is really important while statistical biases always exist. We do make it a priority to collect good data.
Speaker 1: So quick summary of kind of our key points around methodology. First, we make sure that it's specific to all students majoring or minoring in computer related a computer related discipline. Speaker 1: And like we mentioned already all survey responses are fully confidential and participation is voluntary. Speaker 1: And then you might see this in some of our graphs we show later.
Speaker 1: But to prevent any risk of identifying participants, any demographic groups that make up less than 1% of the survey population are excluded from the aggregate analysis. Speaker 1: Our survey data is filtered with spam detection and then only aggregate data. So we don't display school specific data is released as a benchmark. Speaker 1: And then just kind of our overall metrics.
Speaker 1: We've run six campaign cycles so far and have surveyed at a total of 24 college campuses. So across those six campaign cycles and 20 plus college campuses, we've had a total of 500 or 5,000 sorry 5,545 survey participants and over 700 portrait photo participants. And then in terms of our data, we survey across a lot of different demographic groups.
So we kind of analyze across groups like race, ethnicity, gender, sexuality, first generation, students, low income, and disability. So first we'll kind of look a bit closer at some questions specific to student culture. Speaker 1: So one really interesting question we ask is someone has once claimed to me that a blank has unfairly contributed to my acceptance to my university.
Under this question. We have a couple options for which participants can select multiple such as my race, my gender, my ethnicity, to kind of fill in that blank. In this kind of slide, we've included the analysis across race. But if you go to our website, you can check out the analysis across each demographic if you want to.
This is an important question in particular, because it's so integral to not only feeling that you belong, but also feeling that you deserve to be studying and learning at your school. Here you can tell there's about a 50% difference where at the highest 61.6% of black or African American students have been told their race has unfairly contributed to their acceptance at their university versus kind of on the lowest end only 10.6% of white students have been told this.
Speaker 1: So we can move on to another question for the audience. So this is the same question or sorry, similar question. This doesn't have to do with acceptance to university, but actually in gaining job opportunities. So we'll wait a few minutes to see. Speaker 1: Speaker 2: time similar question to before but specific to job opportunities instead of acceptance to university and then here we're breaking it down by gender instead of race which is what we looked at last time so yeah again kind of a really important question especially as college students prepare to enter the workforce these are the people you might recruit the people you might work with your future co-workers and here we actually see about a 40 difference where 55 of students who identify
as women have been told that their gender has unfairly given them an advantage in gaining job opportunities as well as 38 of those who identify as non-binary or third gender so this is actually a story from that i heard from steph but she said that a funny thing is she once told the recruiter about this stat and that the recruiter was actually shocked and double checked with her that she hadn't misspoken and that the stats weren't the other way around assuming that underrepresented students would be told less that they had an unfair advantage compared to groups with a lot of representation so it really is interesting to reflect on kind of successes how much of that that is earned who truly has the advantage here and what classifies as unfair or fair
another question on student culture relates to microaggressions these are comments that the people around you your peers your friends what they say to you and have a direct effect on your sense of belonging in your groups and communities the graph i've included analyzes across transgender versus cisgender identifying students and we see a 25 gap with 54 of transgender students saying that someone has commented on their identity in a manner that made them feel uncomfortable angry or insecure so it is important to think about how these comments affect our sense of belonging in a group how how can they affect how we look at ourselves and how we compare to our peers how can this change how we interact with others in and outside of the classroom so we do survey out a couple other questions similar to this
about how kind of students feel among their peers and yeah just climate and student belonging cool so now moving on to a few questions more specific to the learning environment kind of speaking of the classroom it's also critical that we survey about authority environment at the end of the day students go to college for education to learn to expand their horizons so it's really important to understand how effective the current education system is and how its design affects people of different backgrounds so one interesting question we ask is whether students generally feel comfortable asking questions during lecture um and people check if they agree or disagree with this statement uh and this uh analysis here specifically for gender identity um but we also have it across all the other demographics if you check
out our website so you can check that out on your own time um so you can see here there's about a 20 percent difference with men at um 51.6 percent and women at 28 percent non-binary students at 34 percent um but it is surprising even that uh male students it's only about 50 percent which is not very high considering lectures the main mode of education in college these days um so that's another thing to look at um but yeah you can see there's a pretty big difference um and we kind of it does bring up questions about whether this education system is actually designed to enable effective learning for everyone how can we maybe think about um making this more more equitable and everyone feeling more comfortable participating
and then another question we ask kind of about learning um learning environment and this is a pretty important one is how intimidated people feel studying computer science and related fields computer science and i definitely felt this personally can be an intimidating field and the intimidation factor may not only affect academic performance but also who enters the major and retention who stays in through the four years this graph specifically is uh broken down by low-income students versus non-low-income students and we see about almost a 10 percent difference where 51 percent so slightly over half of low-income students feel intimidated studying computer science and related fields um computer science can be a really intimidating field um the median annual wage for information technology was 86 320 in 2018 which is 36 about 36 000 higher than
the median all other occupations and especially as the field continues to evolve how can we how can these discrepancies in particular widen the already growing socioeconomic gap in our country so you can see from so much of this data there are noticeable effects and indicators of the student culture and learning environment from across all kinds of dimensions again i wasn't able to cover every dimension but i hope this information was thought provoking and inspired it inspires you to check out the rest of the data um and this is kind of all our share all our all i will share on um kind of specific demographic data but we do have a few more i want to share on kind of summary data so things we ask um across uh kind of and don't break it down by demographic um
um so the next one we'll look at some trends over time and this is actually interesting to look at because now that we've done six surveys we can see there are some trends over time across our survey population um so this data this question in particular is asking about whether people believe diversity organizations whose purpose is to support underrepresented or marginalized groups such as but not limited to the national society of black engineers society of hispanic professional engineers and women in computer science are still needed today so as we can see in 2020 on the left um 80.7 percent of our respondents agreed with this statement and three years later when we just surveyed this past year or sorry this year um that number has actually increased by 6.5 percent to 87.2 percent of respondents
agreeing with this statement the percentage of those who disagree has fallen by 2.5 percent while the percentage of respondents who neither agree nor disagree has fallen by four percent so we're continuing to work with departments to provide analysis of data trends over time to also help evaluate the effectiveness of new diversity initiatives and programming and their impact the culture and how their they impact the culture on their campuses so kind of now that we've gone through all this data we definitely want people to reflect and think about i think this is something we think about a lot um is where this data where this data actually fits into your life um some things you can think about are do these survey results um about college climate surprise you are similar results reflected in your workplace um does your
current workplace have any uh internal climate surveys and what impact if any do they have um can you use these results and bring this back to your community do you consider volunteering with high school or college students and what impact can you have in changing the narrative here um all right and then of course like i said already multiple times um we can only uh present a small portion of our data and we have all of this on our website um so we'd love for you to go check out more of our data on our website um we have more specific interactive dashboards as well as a case study on women in stem uh so you can go to percentageproject.org data to see that awesome thanks luna so uh we've gone through the data and once we've seen that data and thought
about it and reflected how it fits into our life and our communities how can we actually put this data into action so i'll go over what it what is data driven advocacy what are we actually talking about here uh discuss how our organization the presented project achieves this as a case study some future steps and also how you can get involved so without further ado what is data driven advocacy so to us data driven advocacy can be broken down into three main steps the first step is to inform and this is a crucial first step once we have data how do we actually effectively present it to others and create awareness this involves you know translating digesting the data in a way that is actually easy for others to learn uh to reason and to think about
the next step is to help others actually reflect so how do we provide platforms for people to think critically about the data that we're we're producing it's important to really allow for conversation reflection so that people can interpret and decide for themselves what to think about the data that we're giving them and the final step is advocating and this means using data to make real change so once we have informed encourage reflection how do we actually create action so now that we've talked about what data driven advocacy is let's get back to our mission and our goal that we've established as an organization so as you may remember from our earlier conversation we talked about our three-pronged approach as a non-profit we defined three groups that we really hope
to empower and to collaborate with and at the end of the day to impact so as a case study let me show you about how we've applied data driven advocacy so everyone here likes matrices right we're all women in tech so to think about data driven advocacy we kind of use this matrix framework where the columns are steps that we're taking and the rows are the audiences that we hope to advocate for and with so the blue cells are now prompts for us to actually brainstorm and explore actions that we can take so in general it's easiest to start from the top left corner and later expand outwards to further steps and to other audiences so for example let's look at cell 1a it's to inform
students and as a case study one of our main challenges when we started the project was figuring out okay now that we've collected all this data how do we actually get the word out to other students we decided to organize a community photo gallery where everyone could sign up for photo shoots and display statistic that they connected most with so if we go to the next slide so as you can see by putting these spaces to the statistics and showing these stories behind the numbers we found that students felt more connected to the message and we're also excited to learn more so from there creating awareness became more organic grassroots and expanding largely through word of mouth the next step in data-driven advocacy is to reflect on the data and get a deeper understanding of what it
says about the organization and its climate so as an example uh students and departments can reflect on data in an event that we call a diversity summit at a diversity summit such as the one that was held in university pennsylvania in recent years is an event where members of the academic department so faculty student staff and students can come together and reflect on the data that their climate survey has produced so they can hear about diversity related issues firsthand and then actually work together to identify ways to address those issues and for to form a more inclusive and supportive community and then there's a bit of a waterfall effect here right so resulting from the events such as diversity summit departments can actually then further inform the academic community and school administrators
through data dashboards and uh and data reports and they can advocate and plan for further change to department uh through the department of through a divert departmental diversity road map we've also noticed that action in one area can grow and take on new goals and audiences so in our case students from other schools started noticing their friends sharing photos from their gallery and asked if they could actually bring the project to their university and that's when we started to actually think about the realize that this could grow um that's when we launched a social media campaign that serves to inform the community and power and that is powered by students so this social media campaign is facilitated via our campus directors at partner schools so on the same day every year
everyone that participates in the campaign across all the different schools that we work with share their photos and advocate for underrepresented students in computer science it's been a really exciting opportunity for students to share their own stories and their own experiences and how that relates to the data and it also takes advantage of the power of social media to get the data out there in a consumable and objective form so over the past four years by informing students and faculty and empowering empowering them to take action we provided a platform where they can advocate and inform others and we see this as fuel fast-paced growth in the number of participating schools photos and survey respondents each year we can see here how the number of schools has grown the photos
has grown as well as the survey respondents over time as well even though our survey methodology changed in the middle of it so lastly to encourage the community to reflect on data we actually share you can see the 3c column we actually share our data our aggregate data and hold an opportunity on our website for people to look at our interactive dashboard which basically has intriguing data analyses and visualizations we publish a visualization of our average data across all of our schools on our website and that serves as a public benchmark benchmark that can advocate and raise the bar as more schools join the initiative so uh without further ado gonna talk a little bit more about what we're working on now yeah so now um we are focusing on growing school participation and expanding internationally
we're expanding on data metrics and visualization as you can see from our website and leveraging a community through a datathon we're expanding the program beyond the campaign with more actionable follow-ups to the data um like with things with events like the diversity summit for instance repeat and then we're increasing data visibility among students faculty and school administrators and also expanding partnerships with other non-profit organizations and hopefully raising awareness um about the work we're doing at events like this and then in addition to what we're working on now um we always have opportunities to get involved um you can join our newsletter follow us on social media um and we're always opening up new positions and looking for more volunteers um we actually have a bunch of slots open right now on
a number of different teams uh so you can check out our link tree there if you're interested in volunteering um if you'd like to support us through donations you can also feel free to go to our website at percentageproject.org donate or you can also buy our merch with death is wearing right now one one version of our merch and all proceeds for that go to the non-profit and funding our programs and research so so that's it thank you very much and definitely check us out on our website and other social media platforms let me open it up to questions now if there are any which you can use the slido for or the chat Thank you.
I think we have one question. Are there ways to analyze the data at intersections of various demographic groups? Sure. I think as we move forward, we have a wealth of data. And as you can see, we're really collecting a lot of different questions across a lot of departments across the country and a lot of different students who are responding to our statistics and our survey.
And so it becomes really important to actually leverage all this data and analyze it in as many different possible ways as we can. And one of those ways is trying to look a little bit more at intersectionality and understanding. So as you can see here with our demographics, we're looking at a specific demographic and the different categories within that.
But then understanding how different intersections of those demographics and how the data looks there is definitely an interesting area that we are looking to do in the next coming year So hopefully, as you stay tuned with us, we will be posting most of the information on our website as well as Instagram. Yeah, I think that concludes our Q&A section.
So thank you so much for stepping in and taking a look at our session today. I know we presented a lot of data. We presented a lot of different actionable items that we can take from that data. And we hope that you feel inspired and intrigued to take a look at our website or to get more involved through either volunteering or even applying this data and the learnings that you have to your own life and communities.
Thank you again for joining us and hope to see you soon.
Transcribed automatically; names checked against the speaker record. Select a time to jump there.
Article based on this talk
How student climate data can turn belonging gaps in computer science into change
Stephanie Tang and The Percentage Project show what surveys of thousands of computer science students reveal, and a three-step model for acting on it.
In short
Stephanie Tang, CEO of The Percentage Project, argues that measuring how students feel, not only who they are, is the starting point for closing gaps in computer science. Together with co-presenter Luna, the nonprofit's chief outreach officer, she walks through survey findings from students across more than 20 campuses and a three-step model (inform, reflect and advocate) for turning those numbers into concrete action.
Key takeaways
Embracing diversity and feeling a sense of belonging
The power of words
Inform, reflect and advocate
Diversity and belonging are two sides of one coin
Tang opens with a principle about data itself: before presenting any numbers, she discloses who the team is and why it collects them. "It's important for you as a consumer of data to know where your data is coming from." (0:47)
The Percentage Project's mission is to help people of every background embrace their diversity and feel a sense of belonging in the field they pursue. Tang acknowledges that celebrating difference and bringing people together can look contradictory. "Some might argue that these feel like opposing forces, but I want to make it clear that these are not mutually exclusive." (1:56) For the team, belonging comes from understanding people's backgrounds, not from pretending not to see color or gender.
The organization works along three prongs. Students are informed through climate surveys run with student leaders each year. Departments receive individualized reports and dashboards to guide data-driven decisions. The wider community gets an aggregate benchmark that each school can compare itself against, which drives accountability in technology education.
Polls that surface hidden gaps
The session turns interactive with live Slido polls. Attendees answer whether they sometimes believe their success has been the result of some kind of error, then see how the project's survey respondents answered. Across 1,500 students in the past survey year, as the speakers cited, about 53% said rarely or never and about 24% said often or always. Broken down by gender, 18% of men, 27% of women and 44% of non-binary students said often or always.
A second poll asks whether people compare themselves to others and assume those around them are more intelligent. In the survey, about 60.9% said often or always, rising from 50.59% of men to 69.27% of women and 76% of non-binary students.
Luna explains the methodology behind these figures. "When it comes to data collection methodology is really important while statistical biases always exist. We do make it a priority to collect good data." (11:11) Participation is voluntary and confidential, groups below 1% of the population are excluded to avoid identifying anyone, responses are spam filtered and only aggregate data is published. Across six campaign cycles the project has surveyed 5,545 participants at 24 campuses, with more than 700 portrait photo participants.
The power of words
Some of the starkest findings concern what students are told by others. Asked whether someone has claimed that their race unfairly contributed to their university acceptance, 61.6% of Black or African American students said yes, compared with 10.6% of white students, as the speakers cited. "This is an important question in particular, because it's so integral to not only feeling that you belong, but also feeling that you deserve to be studying and learning at your school." (14:03)
The same pattern appears in the job market: 55% of women and 38% of non-binary or third gender students report being told their gender unfairly gave them an advantage in gaining job opportunities. Luna shares a story from Tang, who once mentioned this statistic to a recruiter. The recruiter was shocked and double checked that the numbers were not the other way around.
Microaggressions show a 25 point gap, with 54% of transgender students saying someone commented on their identity in a way that made them feel uncomfortable, angry or insecure. In the classroom, 51.6% of men but only 28% of women and 34% of non-binary students feel comfortable asking questions during lectures, and 51% of low-income students feel intimidated studying computer science. Over time, agreement that diversity organizations are still needed rose from 80.7% in 2020 to 87.2%.
The speakers invite viewers to reflect: do these results surprise you, are they mirrored in your workplace, and does your workplace run climate surveys that lead to change?
From data to action: inform, reflect, advocate
Tang then defines data-driven advocacy in three steps. Informing means presenting data so others can digest and reason about it. Reflecting means creating platforms where people interpret the data for themselves. Advocating means using it to make real change. She maps these steps against the three audiences in a matrix and recommends starting with informing students, then expanding outward.
The case study shows each cell in practice. A community photo gallery, where students posed with the statistic they connected with most, made the numbers personal and spread by word of mouth. Diversity summits, such as one held at the University of Pennsylvania, bring faculty, staff and students together to reflect on their department's data, which can lead to departmental diversity roadmaps. When students at other schools saw the photos, they asked to join, which grew into an annual social media campaign run by campus directors on the same day each year. "We believe that we are stronger together, and our focus is on connecting campuses from all over to advocate for equity in computer science." (3:43)
Looking ahead, the team is growing school participation, expanding internationally, running a datathon and building partnerships with other nonprofits. In the Q&A, Tang notes that analyzing the data at intersections of demographic groups is an area the team plans to explore next.
Watch the talk
Exploring Data Driven Advocacy, Stephanie Tang, 38 minutes: https://www.womentech.net/video/exploring-data-driven-advocacy
About the speaker
Stephanie Tang is Founder and CEO of The Percentage Project, a nonprofit dedicated to data-driven advocacy for underrepresented students in computer science, which she co-founded in 2018 and expanded to 22 universities across the United States and Canada. She is also a Software Development Engineer at Amazon Robotics, working on technical challenges in robotic grasping and manipulation.
Questions this talk answers
What is data-driven advocacy?
How does The Percentage Project protect survey participants?
Do imposter feelings differ by gender among computer science students?
How did a campus project grow to many universities?
Can the data be analyzed at intersections of demographic groups?
Drafted with AI from the talk transcript. Editor review pending.
Frameworks and ideas mentioned
Data-driven advocacy: inform, reflect, advocate
Three-pronged approach: students, departments, community
Advocacy matrix of steps by audiences
Resources mentioned
The Percentage Project website data dashboards (percentageproject.org data)
Case study on women in STEM on The Percentage Project website
The Percentage Project newsletter
The Percentage Project Linktree for volunteer positions
percentageproject.org donate
Cite this talk
Stephanie Tang. "Exploring Data Driven Advocacy." WomenTech Network, 2023. https://www.womentech.net/video/exploring-data-driven-advocacy