We all want more diversity in tech. We rarely acknowledge that the experience of inclusion is the product of Org Design. Presented at O'Reilly Design Conference with Molly Beyer, #OReillyDesign, these slides share some practical tips and advice on increasing diversity through applied design thinking. Learn how to empathize and ideate in response to real needs instead of getting people to 'hack a hairdryer'.
Designing for Diversity in Design Orgs (Presentation)
1. Designing for Diversity in Design
Organizations
Presented to: O’Reilly Design
Conference, March 2017
Molly Beyer - @Beyer_Molly
Eli Silva - @EliSymeon
2. Who are We?
Molly Beyer, @beyer_molly
Anthropologist thinking about thick data in
civic tech. Computational ethics, public
health, ebola, emergencies. feminist. Likes
political art, noise, bikes.
Eli Silva, @EliSymeon
Product designer advocating for
#DiversityinTech. Cloud Computing, Design,
Ethics, Mutual Aid. Likes pancakes, politics
and languages.
6. Neurodiversity (n.)
the range of differences in individual brain function and
behavioral traits, regarded as part of normal variation in the
human population; used especially in the context of autistic
spectrum disorders.
7. “...autistic people shouldn’t have to pretend to be “normal,” to
be conventionally social, in much the way African-Americans
should not have to pretend to be white to be professionally
successful. True diversity entails a more radical form of
difference...one more extensive than what he regards as the
accommodational agenda.”
-David Platzer, Dissertation 2017
8.
9. GenderMag.Org
A tool for testing for Gender Bias in software tools developed by Margaret Burnett,
Professor of Computer Science, Oregon State University
10. Existing gender differences in software
Motivation
Information Processing
Computer Self-Efficacy
Risk Aversion
Tinkering
11. “Ignoring or removing gender
[or other intersections] in
users makes it harder to
address implicit stereotypical
assumptions.”
-Ewa Luger, A Design for Life: Recognizing the
Gendered Politics Affecting Product Design,
2014
15. “...eliminate the possibility that we can
function as all-knowing silent
interrogators”
- bell hooks, Teaching to Transgress, 1994
16. “The first job we have to do is not design
for ourselves.”
– Erica Eden, 2015
17. Evaluating Impact
User-test across multiple intersections.
Connect with primary outcome of participatory research.
Are you answering the identified problem?
23. Many Diversity & Inclusion Initiatives are conceived
and executed with little visibility into what the
target audience desires or needs, and what
barriers they actually face.
24. After spending two years and
$265 million on the effort,
Google’s employee population was
only 2% black in 2016, the same
percentage as it was in 2014.
- Beth Winegarner, “Google’s Hardest
Moonshot: Debugging Its Race Problem”
Fast Company
25. “Top universities graduate black and hispanic
computer science and computer engineering
students at twice the rate that leading
technology companies hire them.”
– USA Today Study cited by Bonnie Marcus, Forbes 2015,
“The Lack of Diversity in Tech is a Cultural Issue”
26. Reducing the distance between designers and the
many populations they serve is perhaps the single
most important charge to the design profession
today.
– Beth Tauke, Korydon Smith, and Charles Davis;
Diversity and Design: Understanding the Hidden Consequences
28. Diversity Debt (n.)
A concept in organizational design that illuminates the extra work that must
be done when decisions about culture and diversity are ignored or optimized
for short-term gains instead of applying the best overall solution.
29. - Susan Wu, “Welcome to Diversity Debt: The Crisis that Could Sink Uber”
30. “Some people don’t like to take responsibility for
their own shit. They blame everything in their life
on somebody else.”
—Uber CEO, Travis Kalanick, 2017
31. “If you’re not asking ‘how could this be used to
hurt someone,’ in your design/engineering
process, you’ve failed.”
– Zoe Quinn
34. Literally start with people (Empathize)
Who is telling our inclusion story?
Ask if the story your company tells about its own
diversity and inclusion is coming from the ground up.
Suggestions:
Define ‘Culture Fit’ very tightly, on paper, so there’s
no ambiguity. Revisit often to check for bias.
Look at job descriptions you control for words like
“Dominate, competitive, pleasant.” (Descriptions that
use biased language get 42% fewer submissions.)1
Run language in job ads through a Gender Bias
Reduction Tool (Gender-decoder.katmatfield.com,
DotEveryone.com)
Image Credit: Pexels, Data Source: ZipRecruiter
35. Define the problem
What does the problem actually look like?
Suggestions:
Journey map your diversity gaps for biases and
obstructions.
Focus on high impact, high frequency problems
first, as these are likely to most demonstrably affect
some change.
Pick specific targets within these problems to
address, using a matrix of impact to
employees/candidates and ability to solve.
Generate specific insights and problem definitions.
“We use gendered language in 46% of job listings” is
a lot more powerful than “We have some bias in job
listings.”
36. Ideate in response to REAL needs
How does our idea meet the
needs we have heard?
Suggestions:
Before you generate a single idea, pick a feasible
goal, that you can align on.
Seriously and without question, involve the people
you’re reaching out to. Listen to them and let them
help steer your efforts.
Learn from your own internally marginal groups and
empower them to generate AND implement ideas
for change. (Women, Minorities, LGBT, Disabilities,
Neurodiversity)
Define a way to measure impact before you start
building something.
Image Credit: Pexels
37. Prototype for best fit
How might our culture/process
hurt someone?
Suggestions:
Do your research. Evaluate how your idea to
increase diversity has collected feedback and
gained visibility into the lives of those you’re
designing for.
If your campaign or effort focuses on a marginal
group, bring them to the table during the idea
phase, give them the power to shut you down.
Ask, “How might what we’ve built here hurt
someone,” BEFORE you release it into the wild.
Image Credit: William Stitt
38. Prove it
Did we build it right?
Suggestions:
Make sure you align to the double goals of “have a
positive, measurable impact,” and “specific solutions
to specific problems.”
Build a feedback loop into new D&I efforts, as well
as existing ones, and react to feedback as quickly
as possible.
You only get to succeed if you can demonstrate:
[ ] We Listened.
[ ] We Included.
[ ] We Empowered.
41. “As designers, we find ourselves not just
functioning as human-computer interface
designers, but as designers of an interface
to systems that never saw ‘users’ coming”
– Gretchen Anderson, “Designing for Social Impact”
42. “Washing one's hands of the conflict between
the powerful and the powerless means to side
with the powerful, not to be neutral.”
– Paulo Freire, Pedagogy of the Oppressed, 1968
Inclusivity being the operative term here. I’m not here to convince you that diversity issues exist-- that’s a separate argument that most who study social processes are over. There’s a lot of data out there to support the lack of diversity in tech hiring and in the products it produces. Rather, I’m going to try to talk to you about uncovering hidden bias and avoid designing bias in products or services in the first place. And then Eli is going to talk about how to do this on an organizational level.
By centering your user, a complex human with social need, design is already inherently mindful of diversity or ethics, at least in it’s “pure” or “theoretical form.” However, we don’t always do a good job about thinking about this critically enough in practice.
Source: http://www.metastatic.org/text/This%20is%20Water.pdf
'There are these two young fish swimming along, and they happen to meet an older fish swimming the other way, who nods at them and says, “Morning, boys, how's the water?” And the two young fish swim on for a bit, and then eventually one of them looks over at the other and goes, “What the hell is water?”'
In his now famous commencement speech, David Foster Wallace encourages a graduating class of students to focus on the mundane challenges of everyday life, to recognize they can control HOW they look at the world. Our design and engineering environments, often the whole process of creating experiences is much like these fish who don’t realize they are swimming in water. We often do not recognize what our environments do to contribute to a lack of diversity.
Like for instance, who’s talking about neurodiversity? Except my wonderful my friend David Platzer who does his research on hiring of neurodiverse people in the bay.
Platzer’s ethnographic research on autism. His research participant labeled this the neurodiversity paradigm.
Autistic people don’t want to have to conform to polite social standards and noisy work environments, but we keep training to do them. We should not pathologizing autism, we should be treat it as another form of diversity, in this case neurodiversity and design our workplaces to meet their needs as well, not expect them to bend to ours.
Designers aren’t oracles. Often we create products that are fundamentally not usable by all intersections because we are not all knowing. We reinforce dominant culture in our hiring practices as well. I posit that we cannot be trusted to make ethical decisions without being intentional and having a framework with which to do that with. We will try to cover some methods to do that here.
Discovering ender issues within software has received almost no attention. Margaret Burnett and her team started looking at the ways software currently supports or inhibits male or female problem solvers.
GenderMag team identified common problems already identified by others dealing with gender bias in problem solving softwares such as spreadsheets, visualizations, online classwork platforms, intelligent systems, scripting programs, and web/home appliance dev. This research, while focused on male/female differences, in it’s solution does address gender spectrum issues.
Motivation
Males- technology itself
Females- Motivate to use tech for what it enables them to accomplish
Information Processing Style
Males- selective styles such as following first promising information and then backtracking
Females- comprehensive information processing
Computer Self Efficacy
Females- lower self-efficacy so impacts how willing people are to use “hard to use” features
Risk Aversion
Females- more risk averse in general which affects which feature set they use
Tinkering
Females- less likely to experiment with new features but when they do benefit more than males
Males- some tinker excessively with affects design decisions underlying software that assumes users will learn new features by tinkering with them
The GenderMag method is a set of faceted personas that bring the aforementioned five facets of gender difference research to be used during cognitive walkthroughs
The cognitive walkthrough is a usability evaluation method in which one or more evaluators work through a series of tasks and ask a set of questions from the perspective of the user. Standard CW’s usually don’t use personas, but this paired method can be less costly.
However, the creation of personas requires care to be well researched and actually informed by existing realities of intersections of people, like the GenderMag project. There is a danger to describing users in a very abstract level, or one that isn’t based on actual research of your users. “People who may use existing product” is too narrow! Create the CW specific to your product
So, pairing cognitive walkthroughs with well researched intersectional personas is one way to help create diversity.
Paulo Freire, listening - dialogue - action approach.
If this sounds familiar it’s because we use similar methods in steering products through continuous feedback.
Ongoing interaction between reflection and action people take.
If you’re not already doing this, pairing a researcher throughout your entire design process with a designer and engineer can help to reach towards this goal.
Have your user define your problem to begin with. Go back further with your user than you think, and make exploratory research a priority. This will give the person you’re designing for more participation, and thus power, in the design process. We are designing for *people* with diverse needs different than our own.
Certain kinds of participatory research are cooperative system design, ethnographic field methods, and participatory action research.
Give your research team time to actually explore through ethnography, immersing themselves in the context of your users lives, and define your users across *all* intersections.
This will eliminate the possibility that we can function as all-knowing silent interrogators.
And give us the ability to illuminate actual user needs in a diverse way.
Because it is our job to design for the “other” and that takes knowing them.
Source: https://eyeondesign.aiga.org/we-all-want-more-diversity-in-design-but-what-are-we-doing-about-it/
Intersections = genders, ethnicities, disabilities, etc.
Create metrics for evaluation that are defined as problems. As Julie Zhou said yesterday, don’t derive what your goals are from what you can measure, derive what you should measure from your goals. Let the people you’re designing for, who you as a team recognize as diverse and complex, inform your definition of the problem you discovered in the research phase.
So, are you actually answering the identified problem across diverse users?
Source: https://www.forbes.com/sites/bonniemarcus/2015/08/12/the-lack-of-diversity-in-tech-is-a-cultural-issue/#48b80bfe79a2
This is true for people on the autism spectrum, as well as other intersections.
Make sure to have feedback loops or a way for people to provide you critical feedback loop that are anonymous and accessible!
Hack a Hairdryer
Tay- Twitter Bot