The Magazine of the Usability Professionals' Association
By Neema Moraveji, Rania Ho, David Huynh, and Leizhong Zhang
While there are many problems to be solved in China, we chose to address the need for communication between migrant workers and their children left behind at home. Although these workers lack exposure to computer-based tools, they welcome changes to better their lives and thus, they make willing subjects. Their weaknesses—including a low rate of literacy—amplify their detachment from Western design issues and make them interesting subjects.
These workers’ migration has raised social concerns in China. There are 114 million Chinese migrating to urban centers to find work. Many of these migrant workers are parents who leave their children at home in the villages in the care of the children’s grandparents. From interviews, they confided that their main problem was parenting their children from afar. They believed that better means of communication with their children would help alleviate this problem.
As modernization of China proceeds, few efforts are specifically focused on the migrant worker population. The goal of this prototype is to improve the quality of communication between a migrant worker parent and his or her children while they are separated.
During Phase 1 (April 2005–June 2005), we chose a scenario, interviewed stakeholders, and started to sketch a solution. During Phase 2 (June 2005–October 2005), we continued to identify specific design issues and started conducting user studies.
Previous Relevant Work
Chinese cultural traits that inform user interface design have been researched widely, and the number of studies grows every year as China gains in importance to Western technology companies. These studies, such as the ones documented in Geert Hofstede’s book Cultures and Organizations: Software of the Mind, are largely ethnographic, anthropological, or psychological. Less focus has been on user studies with interactive devices. The inaccessibility of the target users is most likely the reason for this.
As many parties are affected and interested in the phenomenon of urban migration (including local and national governments, local construction companies, and telephone companies, for example), there has been a great deal of sociological and economic research on the migrant worker population in China. These studies are largely aimed at understanding reasons for migration, consequences, eco-nomic factors, and policies that deal with this phenomenon. However, design considerations for communication between migrants and their families back home have not been studied from an interaction design perspective.
An initial interview was conducted with a migrant worker in her thirties who works as a domestic helper in a private home in Beijing. She has two children aged eleven and fourteen, whom she has left back in her village with her in-laws. Her husband is also a migrant worker in Beijing. He lives and works on a construction site.
Remote Parenting: Our interviewee said she spent 90 percent of herphone conversations with her children and only 10 percent with her parents. She made it clear that parenting her children was her primary concern. Therefore, in villages, the primary users of our communication system will be children and young adults.
Synchronization: To contact her children, our interviewee needs to call oneday ahead of the regular market day and ask the shopkeeper in charge of the public phones to inform the children to wait near the public phones at a certain hour on the market day when she plans to call back. Unpredictable schedules often reduce communication between parents and children to one call per month. Minimal parental presence in the children’s lives as a result of such infrequent communication makes parenting difficult.
Literacy: Migrant workers have varying levels of literacy. They are notalways familiar with Pinyin, the Romanization of the Chinese script, which is generally used for input on Chinese desktop computers.
Familiarity of Surroundings: Live-in housekeepers rarely explore the areasurrounding their workplace. However, they do know where the local food market is.
Surveys indicate that China has 39,000 towns in total. Each town is frequented by inhabitants of a dozen surrounding villages for services such as its market, its public telephones, and such. We estimate that each town’s public telephone serves as the rendezvous point of communication for the children of about one hundred families. We estimate about 500 parents share the public phones located near local produce markets.
The most helpful research conducted took advantage of our geographical location. The design process was informed by personal experiences of the group, two stakeholder interviews, and focused user studies.
As the research team is based in Beijing, shared personal experiences from all the researchers helped to increase the general collective knowledge on the migrant worker phenomenon, lifestyles of housekeepers in urban centers, Chinese telecommunication standards, rural phone use, and so on. As it is difficult to gain much of this information during the course of rapid prototyping (since villages can take days to reach), our collective experience proved valuable.
In addition to the interview with a housekeeper, a field visit was paid to an elementary school for migrant workers’ children on the outskirts of Beijing. The school’s dean was interviewed at length. After the interviews were performed, several designs for kiosks were proposed, and focused user studies were performed with female migrant workers who work as custodial staff in the research facility.
Design and Resulting Design Decisions
In order to make communication more accessible and affordable to migrant workers, we decided to design a system of public, shared voice mail kiosks at markets in towns and cities. These kiosks let parents leave messages for their children, and vice versa, in an asynchronous manner. We also augmented the voice mail with videos so as to enrich the com-munication between parents and children, thus increasing parental presence in villages where the children reside. With video, parents can also visually inspect their children and perhaps even their school work, while—we hope—feeling better connected.
By locating the kiosks in local food markets, the system can piggyback onto public newspaper reading displays, already prevalent in China. These displays are well-understood by the population as the source for localized, public information. Rural towns also use the same newspaper display system.
A parent wishing to communicate with his or her children must initiate the first contact by using our system to call home and leave a voice-mail message. To do this, the parent enters the public phone number of the home village (or of the nearest town with public phones) as well as his or her full name. This combination identifies the recipient(s) of the first message, namely the parent’s family at home, and serves as the rendezvous point of communication between the two parties for all future messages.
The combination of the home public telephone number and the parent’s full name grants access to the voice mailbox of that family. No user name or password is required. Such departure from the conventional account-access paradigm sacrifices privacy for simplicity. This design choice is justified by our observations that the Chinese are accustomed to trading privacy for other benefits, such as cost or simplicity of use.
Compared to entering the phone number, entering a Chinese name is a more difficult design challenge. Pinyin is not widely recognized by migrant workers, so Pinyin-based input methods that are otherwise used pervasively on desktop computers cannot be used in this case. We turned to hand-writing recognition as the next most promising technology. Our user studies indicated that a tablet PC interface might be suitable for Chinese name input.
Management of voice mailboxes in our system is simplified by not providing the feature for deleting messages. Instead, each message is stored for a fixed time interval. This design choice also prevents one’s messages from being deleted by someone else. This is crucial for such a system as ours in which accounts are not protected by passwords.
The most unique portions of this work are in its context: the evaluation of technology in the context of this specific user segment, the social context of the family and urban settings, and the leveraging of Chinese cultural cues. The usability studies validated some claims; our next steps included participatory design.
Three small usability tests were performed on each of the twenty participants to validate assumptions about literacy, readability, and input. All participants were female migrant workers who held housekeeping jobs. The participants were not employed in a private residence (as our intended users are), but in a place of business. One factor that may have affected the results was that all the participants
were accustomed to working among computers and other technology devices. They may not have been as intimidated by technology as a domestic worker in a residential scenario would be. None, however, had any experience using a tablet-style PC and stylus.
The first task, A1, was intended to see how a stylus and tablet could best support Chinese character input for this user segment (see illustrations). There is a common Chinese practice of writing ideographs in the air with a fingertip. This technique is used when describing characters to others or remembering the number of strokes in an ideograph. We initially explored the idea of letting users write their names with their fingertips in the air. The resulting strokes would be analyzed by a camera or transferred onto a tablet. But as there was a question about technical feasibility, user study B1 was administered with users entering characters with a stylus. For this task, a commercial tablet PC was used. Participants were asked to write their name in Chinese characters using the stylus. We predicted that participants would be able to enter Chinese characters on the electronic system easily after some initial hesitation. The study revealed that most participants had no problems using the stylus.
Another point in administering task A1 is related to the fact that all the participants were not hesitant to use the device. We acknowledge that this could be because a person who is seen as professionally superior to them was requesting that they perform this task.
We also acknowledge that, in our test, character recognition proved to be less than 50 percent accurate. This may have an effect on future designs.
The next issue was for a user to glance at the display to check the kiosk for new messages. One design displayed the names of hometowns that have recently sent messages atop the kiosk. Another design displayed a map of China with geographical indicators of the origins of new messages. Two tests were performed to verify if users could identify the town nearest to their hometown that had a public phone. In the first task, B1, participants were asked to identify the location of their village on a map of China. In the second task, B2, participants were asked to pick out their hometown on a list of towns. Observations from administering task B1 concluded that fourteen of twenty subjects did not easily find their hometown on a map of China. They relied heavily on text descriptors on the map. However, on task B2, all users were able to quickly find their hometown on a list. Therefore, our design eschews spatial knowledge of hometowns and uses a list format instead.
The immediate next step for this study is to create and test a high-fidelity prototype. First, a kiosk would be installed at a market. Replies to user messages would be stock messages from the system. This would simplify implementation and focus on usability issues. Later, rural kiosks would be installed and the interface would be tested with children.
As the original aim was to uncover usability and design issues by designing for a real scenario, we present issues that will be studied in detail as we move forward with the project.
Issues for Further Research
Scale: A huge population with relatively small number of unique names;extensive settlement networks ranging from very poor to very rich; many local dialects.
Chinese character input: What is the best interface for Chinese characterinput and error correction?
Shared access: How do Chinese privacy and sharing norms apply to communal devices? We have observed that Chinese people have fewer privacy concerns than Westerners. How can this idea of privacy inform design?
Conceptual understanding of data: Can we design interfaces for users toconceptually understand virtual messages? Concepts would include performing operations (save, delete), navigating, keeping mental lists, and understanding the implications.
Designing for Chinese novices: How is designing for Chinese novices different from designing for novices in general? How can we communicate the usefulness of a product before it is actually used?
Literacy: How do designs for character-only readers differ from onesdesigned for users who know Pinyin or even for those who are illiterate? Does the process of designing icons for Chinese culture differ from designing icons for a Western audience since the Chinese lan-guage is derived from pictograms?
Socioeconomic extremes: How should designs for the same culture differ bysocioeconomic level? What are the personas for these levels?
Social hierarchy: What is the best way to design user studies and interfaces without social status affecting outcomes?
In our study, one of the two administrators was a foreigner, and both were full-time employees. The subjects seemed to cooperate without reluctance. We attribute this behavior to the social hierarchy inherent in the Chinese culture: one always submits to the command of one’s superior. One helpful measure was to hold conversation and interviews in a private setting (such as a home or empty hallway). For testing on the “high-tech” tablet PC, it helped to keep explanation of the device (and its implied monetary worth) to a minimum, instructing users to “just write on it as they would a piece of paper.” Participants were not paid or given rewards for their time, and there didn’t seem to be an expectation for such.
There are many possible avenues of exploration related to designing for emerging markets, specifically in China. This paper is part of a larger effort to gain practical and specific insight into the usability of various products in China. The design and implementation of this project has already, and will continue to, uncover specific interaction design experiments that can be carried out with Chinese participants. UX
Reprinted from the DUX 2005 Proceedings by permission of AIGA. DUX 2005 was the final World Usability Day session this year.
Neema Moraveji is an associate researcher at Microsoft Research Asia’s Center for Interaction Design in Beijing, China, where he studies cross-cultural design and interfaces.
David Huynh is a research assistant and PhD candidate at the MIT Computer Science and Artificial Intelligence Laboratory.
Leizhong Zhang is a user experience program manager in the Advanced Technology Center, Microsoft, China. His research interests include information visualization, ambient UI, and online social networking.
Rania Ho received her master’s degree in interactive art from New York University’s Interactive Telecommunications Program. She is a writer and performer with the sketch comedy group, 18 Mighty Mountain Warriors, and currently lives and works in Asia.
Usability Professionals' Association
promoting usability concepts and techniques worldwide
User Experience Magazine is by and about usability professionals, featuring significant and unique articles dealing with the broad field of usability and the user experience.
This article was originally printed in User Experience Magazine, Volume 5, Issue 2, 2006.
© Usability Professionals' Association
Contact UPA at http://www.usabilityprofessionals.org/about_upa/contact_upa.html