Nearly 20 years ago, Bill Gaver and colleagues introduced cultural probes to the HCI community . Since then, HCI has seen widespread use of these designed objects that are meant to provide designers with glimpses into local cultures, to inspire design. To date, this approach has been underutilized in HCI for development/information and communication technologies for development (HCI4D/ICTD). A method that is meant to help us understand local cultures, and that also emphasizes creating relationships with people, seems highly compatible with HCI4D/ICTD researchers, who are frequently not members of the communities they study. Because their motivations come from design, probes embody a different set of sensibilities than do other social science research methods, especially in terms of how they engage their recipients and how the information that emerges from them is analyzed. Fundamentally, probes are meant to "subvert...traditional HCI methods" . These intentions, however, are often lost, and the approach has been widely used in ways that are similar to traditional methods in HCI. I attempted to use probes as Gaver et al. originally intended (detailed in The Presence Project ), because doing so can promote discussions within the HCI4D/ICTD communities about positionality in cross-cultural research and the epistemological constraints implicit in traditional HCI methods.
Here I describe my experience using a probes approach during the early stages of a five-year design project investigating domestic technology use in rural households located in Bungoma County, Kenya. To date, I have visited participants' homes twice; my first time was in June 2016. During this visit, I presented the probe activities—responding to comment cards and taking digital photographs (Figure 1)—to 23 volunteer households. In May 2017, I returned to these households to discuss their responses.
|Figure 1. Comment cards box (left) and digital camera (right).|
Whereas most research approaches minimize the subjectivity inherent in their use, a cultural probes approach requires embracing how the designer's positionality influences their work. I work as an HCI professor and have been conducting research in Bungoma County since 2011. I travel there once or twice a year, staying for two to six weeks at a time, broadly investigating people's interactions with technology. While in the field, I am an outsider, a mzungu (Swahili for "white person"). My skin color, education level, and comparative wealth are clear markers of difference between me and most people in Bungoma. I am aware of, and sensitive to, the power asymmetries underlying my fieldwork and post-fieldwork processes, and am committed to conducting research in a respectful and ethical manner. I am also a native English speaker; my Swahili is rudimentary at best, so I require an interpreter when conducting research. Nightingale Simiyu, a Bungoma resident, has worked as my interpreter and field assistant since my first visit to Western Kenya.
How did I develop my probe activities? My decisions were mostly driven by the requirements of the agency funding my research, the institutional review board (IRB), as well as by contextual factors in Bungoma County. My project is funded by the National Science Foundation (NSF), a U.S. government agency that generally values typical science- and engineering-based research approaches rather than the artist-designer ones embodied in probes. My funding was contingent on receiving ethics approval from my university's IRB, and from Kenya's National Commission for Science, Technology, and Innovation (NACOSTI). Gaver does not mention this condition in his writings on probes. Engaging in this approval process prompted me to choose probe activities that I could explain in terms of other research methods used in prior research (e.g., diary studies and photo elicitation). Contextual factors also affected my decisions. Although Kenya has a national postal service (Posta Kenya), mail delivery can be unreliable in Bungoma, and households that are far from town rarely have mailboxes. Waiting for returns to be mailed back was not feasible. Rather than using postcards, I developed the comment card activity. Similar to postcards, this activity provides a mundane medium to ask a wide range of questions . Reflecting on these ethical and practical constraints underlying my project, as well as on my assumptions about which activities participants might engage with, I developed two probe activities: responding to comment cards and taking digital photographs.
I gave respondents a cardboard ballot box with a side pocket. In each pocket were 16 5x4 cards with questions on them (Figure 1). The questions on the cards were inspired by those asked in prior studies—open-ended, slightly provocative, and intended to encourage surprising responses. Questions included: "What is something Americans should know about Kenyans?" "Be quiet and listen. What do you hear?" and "Imagine that you could travel anywhere today. Where would you go and why?"
The second activity was taking photographs with a compact digital camera. I attached a laminated card to each camera's lanyard. Each card contained open-ended and ambiguous prompts, including: "work," "fun," "faith," and "anything you like" (Figure 1). Cameras were placed in a resealable plastic bag to make them seem like a package.
My research assistant relied on her personal contacts to identify probe recipients. Like Gaver's method, our approach did not attempt to control for demographics; respondents came from a range of circumstances. In June 2016, we introduced ourselves to participants and explained our project. We also obtained informed consent for the study. We added that collected probe returns would be anonymous. Our consent process also included asking for permission to use anonymized versions of their returns in publications associated with this project. All participants agreed to these conditions and verbally consented to be in the study. We then gave participants the box and asked them to respond to one or two cards daily. They were also encouraged to take pictures that corresponded to prompts on the card attached to the camera. We ended sessions by telling them that Nightingale would return in two weeks to collect the comment cards, and that I would return in about a year to talk about their responses.
The questions on the cards were open ended, slightly provocative, and intended to encourage surprising responses.
In May 2017, I returned to the households. All invited me back into their homes. Prior to talking to participants about their responses, I explained my project's progress and again obtained informed consent from them. I then showed them 15 to 30 of their photographs, about which they spoke openly. I received no reactions from participants suggesting they had negative feelings about engaging in the probe activities. In fact, at the end of the sessions, most asked when I would return to their homes to continue our conversations.
Boehner et al. observe that prior research rarely documents what takes place between reviewing probe returns and developing design proposals . Rather than provide a comprehensive analysis of participants' 282 comment cards and 686 digital photographs, I give my interpretations of a few of them here. These interpretations are influenced by my stance, research interests, and idiosyncratic curiosities.
Sensations inBungoma. Responses drew attention to sensations that make up everyday life in Bungoma, including the auditory ones: the rumbling of the 18-wheeler trucks on the Trans-African Highway just outside of town, the sputtering of the piki-piki motorcycle taxis, the radios broadcasting news about Kenya's then upcoming elections, and the song of the African golden weavers (which has been transcribed as pew...pew...tew, chinkichi-chewchew). These and the other sounds (e.g., "frogs croaking") are "amazing" (Figure 2). As are the visual sensations around Bungoma—including the awe that comes when encountering the gigantic boulders that dot the landscape and the clouds in the area's vast sky; indeed, they are "nice" (Figure 3). Responses also bring to mind the smells in Bungoma: the faint odor of manure in people's homes (a result of covering houses with cow dung to prevent termites); paraffin wax burning; the scent of freshly cut sugarcane being hauled to one of the nearby sugar factories; the salty smell that accompanies drying fish; as well as the mustiness that typically follows the torrential downpours that seemed to occur daily in May and June.
|Figure 2. "I kept quiet for sometime, it was at night and I heard frogs croaking everywhere loudly. I heard then some footstep of a dog called mulika mwiza barking. It was amazing."|
|Figure 3. "Those are just clouds, I took this because they are nice."|
Households. Photos captured moments outside and inside participants' homes. Some showed a father and son completing a new concrete cover for the borehole adjacent to their home; others caught newly trimmed hedges outside of someone's small, rectangular, mud-thatch house. Other photos documented families gathered around tables in their home's sitting room eating ugali (maize porridge), sukuma wiki (collard greens), and, less frequently, beef or chicken stew. More intimate moments were also depicted, including people waking from sleep and mothers bathing children. Comment card responses to my question "What do you enjoy most about your home?" mostly reflected a deep appreciation for these spaces: "I enjoy the privacy, the freshness of my garden, and the landlord-free life. And I like to sit with [my] family to eat together and look at my crops in the shamba."
Using cultural probes draws attention to the messiness underlying all research and design endeavors.
Surprises. Probe responses should be surprising, and many of mine were. Some made me laugh, prompting more questions than answers, for instance, the response to "Imagine that you can travel anywhere today. Where would you go and why?" The response in Figure 4 was unanticipated, given that Kenya is home to elephants. Another card mentioned traveling to America for its good climate, a comment I found surprising because I live in a part of the U.S. known for brutally cold winters. The cards also captured the surprises in our respondents' everyday lives— see the story in Figure 5.
|Figure 4. "I would like to travel to Thailand and see elephants."|
|Figure 5. "As I woke up early I found my money which was lost for a week. I was so happy for the money and I rushed to the restaurant to have a cup of tea which was hot and burnt my tongue—but from that day I would never forget."|
Gaver writes that design ideas emerge slowly over time; I also find this to be the case. I am developing a collection of design workbooks to illustrate early design ideas inspired by my responses. In June 2019, I returned to Bungoma to share these workbooks with the 23 probe recipients; I am currently writing about my findings.
My use of cultural probes succeeded beyond my expectations. The hundreds of comment cards and digital photographs from households, as well as the conversations they initiated, draw attention to the diversity, idiosyncrasies, complexity, and—most significantly—the texture of everyday life in Bungoma. They capture residents' multifaceted lives—as well as the sounds and smells that are unique to where they live. In my experience, these insights, as well as non-generalizable, one-off observations, can be difficult to uncover and to report using traditional HCI methods.
The value of reflexivity has been widely accepted in anthropology. Within HCI, there has been a shift among researchers to devote space in their papers to considerations of how design is shaped by their race/ethnicity, gender, nationality, and so on. This is a welcome change that contributes to greater transparency about the relationships between researchers and participants. A probes approach emphasizes how these demographic factors affect research, and how other significant but downplayed factors also affect research and design. In this sense, using cultural probes draws even more attention to the messiness underlying all research and design endeavors, but that is often omitted from HCI publications.
When developing the method, Gaver and his collaborators wanted to challenge stereotypes of older people as "frail and needy," characterizations they argue ultimately lead to designs that emphasize "only the negative aspects of their lives" . My motivation for using them is similar; I want to reject characterizations of developing regions and the people living in them as poor, marginalized, and vulnerable. These are not the perceptions that emerged from participants' probe responses. My hope is that greater use of cultural probes in HCI4D/ICTD will lead to designs that do not narrowly focus on socioeconomic development (e.g., health, education, and livelihood).
Here I offer HCI researchers and practitioners a case study that demonstrates how to use cultural probes. This is not intended to be a how-to guide for using the method in developing regions. Probes were never intended to be a reproducible method, and there is no one right way to use them. My hope is that embracing them can inject more creativity into how research is conducted in developing regions, as well as promote challenging discussions about the nature of HCI4D/ICTD research and how it is conducted.
Susan Wyche is an associate professor in the Department of Media and Information at Michigan State University. Her research focuses on human-computer interaction (HCI) and information and communication technologies and development (ICTD). Wyche received her Ph.D. in human-centered computing from Georgia Tech. firstname.lastname@example.org
©2020 ACM 1072-5520/20/01 $15.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2020 ACM, Inc.