Encoded Behaviors: Gendered cognitive labor patterns in home assistant technologies
Master of Design Thesis — Research Phase
My research is a design inquiry that sits at the intersection of home assistant technologies and cognitive labor in the home. It explores the relationship between home assistant technologies — like Alexa, Google Home, and Siri — their promise of domestic productivity, and how this technology reinforces embedded gendered practices of domestic work.
My cultural background and life experience has played a large role in why these intersections of gendered home labor, domesticity, and technology fascinate me. I was raised in a diverse suburb of Southern California, where the ethnic make-up in my neighborhood growing up was majority Mexican, Vietnamese, and Korean. A large portion of my friends and peers had foreign-born parents. I myself am a second-generation American, after my parents, who immigrated to the the United States as a result of the Cambodian genocide under the Khmer Rouge. It wasn’t until I entered university, in a wealthy, coastal California city, studying in classrooms where I was the minority, that I knowingly felt the impacts and nuances of race and class. In the past, I have worked as a personal assistant to the matriarch of a high-powered music executive, where the domestic help consisted not only of myself, the personal assistant, but also a nanny and a housemaid. I worked for several years in Silicon Valley for a private wealth management firm as an Executive Assistant where I supported several male, Partner-level executives. I also previously served in an operational and administrative capacity at a global design consultancy. I am currently a mother to two children under the age of five, married to a supportive partner with a demanding work schedule, and finishing up my final year of graduate school while balancing two part-time jobs.
These experiences have grounded my views on the underappreciation of domestic work. I have witnessed and felt the amount of physical, cognitive, and emotional care it takes to keep families running smoothly and to keep a company’s operations humming along. Who is going to remember to research and decide on a birthday gift for their mother or mother-in-law? Who is going to manage a calendar so everyone in a household or team has all the details at hand? This kind of work, called cognitive labor, is something that is difficult to quantify and only compounds as one enters different life phases. It often goes unnoticed — it is often invisible — and it is this subject matter that is the underlying basis of my thesis exploration.
For the past year, I’ve worked as a research assistant and design researcher in the home IoT space. As I have become increasingly engaged and familiar with the impacts of these devices in our daily lives as it relates to data, privacy, and data translation, I have started to question the product marketing promises of enhanced productivity and human efficiency — the capacity of these devices to not only entertain you, provide you with statistics, keep your home safe, be your friend — but also serve as a pathway to domestic harmony, where tackling home labor is easily accomplished by way of machine commandment.
The combination of cognitive labor in the home, in a society now where both men and women are employed, in concert with the expanding technological systems in our daily lives, I hypothesize, generates more work for a household, and the assumptions around who is doing the work still falls on gendered lines, where women are still responsible for and complete a majority of the domestic work.
My thesis topic sets out to investigate and reflect on: the gendered cognitive labor dynamic in the home, how these everyday embedded objects reproduce this gendered dynamic, how we can make home assistant technology more useful in aiding one’s cognitive load, and how designers might tackle ways to make cognitive labor visible and tangible. I intend to experiment and create a set of design responses that examines, critiques, questions, and helps make visible the nuances of these interwoven topics.
This quarter, I conducted primary and secondary research, designed a research probe and week-long diary study, held interviews with households and one industry expert, and began to gather themes and synthesize findings from my work, which I outline below.
But first, what is cognitive labor?
Cognitive labor is the thinking aspect of labor, which consists of managing household tasks. This is separate and distinct from physical labor which consists of the physical aspect of helping complete household tasks. It is sometimes referred to as mental or invisible labor, and involves four components, where one has to: 1. Anticipate needs, 2. Identify options for fulfilling those needs, 3. Decide among those identified options, 4. And monitor those needs, making sure decisions are executed and addressed. Allison Daminger, a sociologist and Ph.D. candidate in Sociology and Social Policy, concludes in her research findings that cognitive labor is gendered, and that women continue to bear the majority of domestic housework, both cognitive and physical.
Below is an example that breaks down the physical and cognitive tasks of planning a birthday party.
Existing research on household labor centers on the physical aspect of domestic labor. However, as you see in the example above, the cognitive labor portion accounts for a majority of the work it takes to plan an event. By not reporting this form of domestic work and given the historical exclusion of this data from research, it is difficult to generate a full account of the levels of gender inequality in our homes.
Furthermore, Daminger concludes that without raising awareness of this facet of labor, gender operates as an invisible mechanism that shapes home behavior patterns and thereby reinforces this household gender gap. If cognitive labor remains invisible to people in their everyday lives, this collective, subconscious practice of gendered home labor will persist and these daily routines of patterned behavior will stay unchanged.
The evolution of the domestic technologies and household labor
The evolution of home technologies advanced after World War II when post-industrial America saw a societal shift in which families, mainly men, entered the labor market, and where women stayed home. This generation saw a rise in upward mobility, a move to suburban areas, and the paving of the way for home technologies that were marketed to make home life and housework easier. The technologies, such as microwaves, stoves, and ovens, were built with the assumption that a housewife would be staying home to operate them and did not account for women entering the workplace.
Since 1970, the number of two-parent households where both parents work full-time has increased from 31% to 47%. Currently, over 70% of American women with children under the age of 18 are in the labor force. In a 2015 Pew Research study that analyzed how parents divide their physical domestic labor responsibilities, the data shows that a larger share of household responsibilities where both parents work full-time falls to the mother. Noted in the same study is a piece of data regarding the gender disparity in how mother’s and father’s attribute their division of labor — women are more likely to report that they do more of the household labor than their partner, but fathers are more likely to say they share their responsibilities equally.
“Modern household technology facilitates married women’s workforce participation by not freeing women from household labor but by making it possible for women to maintain decent standards in their homes without…a full-time commitment to housework.”
— Ruth Cowan, More Work for Mother
The technologies produced in the post-industrial 1950’s were also meant to enhance human efficiency, by decreasing the time required for household chores. However, what occurred was an increase in the volume of work, “although the work [was] more productive…and less laborious than it used to be, for most housewives it [was] just as time consuming and just as demanding” (Cowan). With men leaving the home for the office and with kids in school, women of these modern homes were expected to complete the household chores, and the new machinery set certain expectations of order and cleanliness in the home.
The share of housework, despite the uptick of women entering the labor force over the past few decades, and the assumptions on how to best to run a home, still relies on these gendered notions of labor and which still rests primarily on females in the household.
We now have these home assistant devices — the Alexa’s, Echo’s, Google Assistant’s, Siri’s — that continue to trend in popularity, and potential ubiquity, in American homes. In 2019, 35% of US households, or 90 million users, lived with at least one smart speaker and the adoption and sales of these devices hit a surge in growth in 2021. As they become more enmeshed in our everyday lives, I believe it is important to critically probe the parallels between the domestic technologies of yesterday and the digital assistants of today that are becoming more embedded in society, especially because they also illuminate the promise of productivity and efficiency for the home.
I conducted primary and secondary research in order to better understand home assistant technology space, cognitive labor in the context of the home, and the current landscape of product offerings that help tackle the issue of cognitive labor within the home.
In my secondary research, I read academic research papers, books, and articles that centered on: overviews of voice assistant technology; household dimensions of cognitive labor; the rising prominence of voice user interfaces and its how they are used in the home and the potential of its effects on shaping roles and relations in a household; the cultural history of household technology in the United States; the affective, gendered traits and behaviors of digital assistants and voice user interfaces; the digital housekeeping of domestic technologies; and critiques and perspectives on the role that designers and members of the HCI community play in shaping future product offerings of our home IoT technologies.
What I discovered is that there is no comprehensive study that investigates voice user interfaces and how effective they are in tackling cognitive labor. In my literature review, I found that the discussion around gender as it relates to voice assistants focused on the device’s personality, voice, affect. My research aims to expand on these expressions of gender and probe how these devices reproduce gendered dynamics of home labor. Furthermore, most research around the domestic labor of these devices, known as “digital housekeeping” pertains to the upkeep, updating, maintenance, and troubleshooting of the physical devices. Digital housekeeping is now a current of domestic work, and the maintenance of these systems will become increasingly complex as we obtain more technology for the expanding smart home. It is interesting for me to consider the parallels of cognitive work as it relates to digital housekeeping and the cognitive work it actually takes to engage with these devices. Overall, the study of cognitive labor and gendered division of labor in the home is fairly recent, so I was unable to find work from the HCI community that addresses specific topics. I want to bridge the home IoT space with discussions around cognitive labor specifically as it relates to domestic work, the gendered patterns about what kind of domestic work we’re asking these devices to accomplish, and who in the home is still managing and delegating these tasks.
I was curious to know if there are current solution-oriented services that help tackle cognitive load. I found that there is an increasing awareness of this facet of labor and people are creating experiences that are aimed at chipping away at this cognitive weight. For example, Amazon released their latest version of their Echo product line — the Echo Show 15 — which is marketed as a family organization tool. A service called Fairshare is in beta-testing and is a product that helps navigate the cognitive load between two partners in a household through chore sharing reminders and management via an app. Furthermore, a Seattle-based company, Yohana, is a new offering that connects you, via an app, to a network of human assistants that you can offload your home labor tasks to. There is a growing current of products and services that are attempting to solve these issues, directly or indirectly, of cognitive labor.
In order to better understand and evaluate the labor patterns in the home, and to get a better sense of the amount of physical and cognitive labor completed each day, I prepared a task exercise for myself and another participant to complete. We each tracked our physical and home labor over the course of five days and then reflected on the exercise together following completion of the study.
The participant said that, through the study, she and her partner began to say “thank you” more frequently, as a gesture of acknowledgement that they were aware of the other person’s physical and cognitive load. It was through this experiment that I learned that just by making people aware of this form of labor who were unaware of it before, by making it visible and real simply through discussion, was in and of itself an accomplishment for making this topic more broadly understood.
I spoke to an employee at Amazon who works in Product Marketing for the Alexa Team. She mentioned the release of their newest Echo product, the Echo Show 15. It is marketed as a 15.6” smart display that helps your household stay organized. You can use the devices, “to keep the family on track with shared calendars, personal sticky notes, to-do lists, shopping lists, and assigned reminders” (Amazon). She described it as a family organization tool that is meant to chip away at this “cognitive weight” that parents carry. This interview reinforced my belief that cognitive labor is becoming an increasingly relevant and timely topic in American households, as we now have corporations like Amazon directly marketing a home device for helping with housework and family organization.
I recruited for and conducted a week-long diary study with three participants that began with an introduction interview where I asked the participants about their home labor dynamic, their usage habits with their home assistant technologies, and how they use these technologies to help them complete their home labor tasks. The diary study was an iteration of my Research Probe, but in addition to tracking their physical and cognitive labor, I included a daily questionnaire that served as reflection questions on their domestic work for the day. Following the diary study, I conducted exit interviews with questions around notable occurrences and insights during the study.
My research for this was submitted to the University of Washington’s Human Subjects Division and Institutional Review Board (IRB) and was approved for Category 2 — Exempt Status. The purpose of the IRB is to protect the safety, rights, and welfare of research participants and to ensure institutional compliance with federal and state regulations and University policies.
While our plethora of home appliances have assisted us in decreasing our physical output, the increased connectivity to products in our daily practices have, I argue, increased cognitive weight for women, especially now that more women are in the workforce. These devices do not directly address cognitive labor and may in fact increase a person’s cognitive load.
In an interview, one participant reflected, “I don’t think [Alexa] necessarily decreases [my cognitive load], so much that without them, I would definitely forget things a lot…it’s trading out the remember to do this, for the remember to put it in your calendar, and then remember to check your calendar, and then remember to do it still.”
Insights on cognitive labor
Cognitive labor is difficult to measure and does not feel tangible.
Overall, through conversations in my interviews and more informal conversations with friends and peers, I found that cognitive labor can go unnoticed even for the person doing the work — the routines of remember to turn the lights off at night, or making sure that the coffee timer is set for a certain time to brew — those ingrained behaviors and routines, as Daminger writes, is “diffused” and difficult to quantify.
“Generally cognitive stuff feels harder for me. Whenever I have a day where I organize the rooms and dishes, like with physical labor, I can measure. Cognitively, when I tend to someone’s needs, it doesn’t feel like I got anything done.”
The dynamic of home labor is often difficult to discuss
Discussions about cognitive labor in the home aren’t frictionless and a tension can often permeate around this topic. In one conversation and not an uncommon theme, a working mom discussed how when she approaches her family about the topic of home labor, the household members will get defensive about the amount of work they have completed, getting frustrated that it is never enough. One participant, in order to remove the “nagging” or implied tone of her requests, would text her husband instead of verbally sorting out daily chores with him because it didn’t carry any “baggage”. This leads me to a design opportunity for creating an experience or service that is not antagonistic or pointed.
Giving cognitive load a name makes it concrete, understandable and real.
Not many people realize what cognitive labor is, even though everyone does it every single day. I received visceral reactions from people after I would speak about my concept and touch on the complexities of balancing work and home life.
One participant included her husband and son in the diary study and shared with them the concept of cognitive load. Luckily, it helped enable small behavior changes in her family.
“Now that I realize that I was [doing cognitive labor], I can delegate the cognitive load and they are aware how things are done because cognitive load is defined. When they ask what’s for dinner, they know [the work it takes].”
Insights on technology
Management of digital ecosystems and physical devices are forms of cognitive labor
A couple households I interviewed had fairly integrated smart devices in the home. The maintenance of these devices was sometimes an unanticipated chore if a smart device was not integrating properly with another device. For example, a Samsung smart fridge that calendar information would sometimes not integrate with their Google calendar. So not only is there cognitive work in managing schedules through these digital calendars, it is also a cognitive task to manage and maintain the devices that these digital ecosystems exist in. The domestic work of managing these devices will expand as our homes become ‘smarter’ with new devices and technological advancement.
Finicky-ness of tech in tension with its promise
The advertisements for these devices give you a sense that the technology is a seamless instrument in your daily routines. However, in reality, I found that there are limitations to the promises of these devices — oftentimes, features like Alexa routines, which are programmed to trigger a routine or command for a child in the home, would not function as anticipated and would trigger at the time or not trigger at all.
Insights about cognitive labor and technology
Home assistant technology is imperfect but no other viable options
In my interviews, I asked participants why, if their devices didn’t perform as expected and frustrations arose, they continued to use them. One participant responded:
“Because at the end of the day, my phone is already in my pocket. There is already an Alexa. There is already a fridge. They’re convenient and practical. Between already being invested and practicality of access, it’s already there.”
The willingness to experiment with technology even though it is imperfect
There is a curiosity and optimism about technology that leads people to experiment because they believe in the promise of what tech can be. A common theme among my discussions was the wish for an integration of the different mediums into one seamless, cohesive platform — a majority of participants used Google Calendar to manage their family and social calendars and Alexa to create shopping and to-do lists. One participant wanted to create a customized screen display for their kitchen that would display all the information they needed to keep their lives in order.
The design opportunities below are drawn from the insights I gathered over the course of my research and they are:
—Design an experience that measures cognitive weight
—Design an experience that tracks cognitive labor
—An experience that makes cognitive labor visible between couples or people in a household
—Design an experience or artefact using tech as the medium to uncover cognitive labor
—Design an experience or product that externalizes cognitive labor
—Design a new service that fulfills promise of cognitive efficiency
—Design a speculative home assistant device that better anticipates one’s cognitive needs
In the next quarter I will conduct design sprints aimed at generating design concepts for my final thesis work based on the design opportunities outlined above.
Some anecdotes from what I learned this quarter:
Starting is the hardest part, but it’s better to just begin.
Whether it was tackling the stack of printed research papers on my desk, beginning to synthesize my findings, brainstorming and sketching out ideas for design solutions, or beginning to write my research paper, getting over the hump of starting was sometimes difficult for me. Delaying the inevitable only compounded the unneeded stress, so tackling the checklist in smaller increments — like setting a goal to write just one-page of your research paper by end of day or making it a point to set aside 30 minutes to quickly sketch out design solutions — was a less intimidating way for me to complete tasks I knew had to get done.
Recruitment for a study can be a chore.
I didn’t anticipate this but I probably spent one full working day (and more) trying different avenues of recruitment for my diary study. I was also spammed which made things even less awesome. Even when you have recruited and even when an individual has agreed to participate, their participation isn’t guaranteed — so I would cast a strategic but wide net for recruitment and reach out to those in your network who you trust to broaden your reach.
The research process is nuanced and messy, not linear.
I gathered so many nuggets of information during the research process that were not directly tied to my topic but within the space of home IoT. When I read academic papers from certain conferences, there is a cadence and format to each paper, and you begin to recognize the structural patterns of writing across the same spectrum of work. Everything is told in a concise, linear pattern, but the research process is in no way linear — the things you learn become a web of knowledge that form clusters and expands, and expressing and translating every piece of knowledge gained along the way can be difficult to do.
Self-motivation is key in graduate school.
When I began this program two and a half years ago, it had been over a decade since I’d completed my undergraduate studies. I was unsure how I would handle and juggle the responsibilities of being a full-time student with a young family at home. My graduate program provides a rich offering to its students — via the range of coursework offered that is not limited to your own department, through the supportive faculty, and the bright minds of the student population. While the program is challenging and there are milestones you need to reach, what I learned is that, as a graduate student, it is only as challenging and rewarding as you want the experience to be. You set your own bar for how much you aim to accomplish with your thesis and overall graduate experience — faculty will not micro-manage and breathe down your throat trying to keep you on track — staying self-motivated, staying focused, keeping yourself organized, and doing the good, hard work, is up to you.