Culture, Interface Design, and Design Methods for Mobile Devices

Abstract

Aesthetic differences and similarities among cultures are obviously one of the very important issues in cultural design. However, ever since products became knowledge-supporting tools, the visible elements of products have become more universal so that the invisible parts of products such as interface and interaction are getting more important. Therefore, the cultural design should be extended to the invisible elements of culture like people’s conceptual models beyond material and phenomenal culture. This chapter aims to explain how we address the invisible cultural elements in interface design and design methods by exploring the users’ cognitive styles and communication patterns in different cultures. Regarding cultural interface design, we examined users’ conceptual models while interacting with mobile phone and website interfaces, and observed cultural difference in performing tasks and viewing patterns, which appeared to agree with cultural cognitive styles known as Holistic thoughts vs. Analytic thoughts. Regarding design methods for culture, we explored how to localize design methods such as focus group interview and generative session for specific cultural groups, and the results of comparative experiments revealed cultural difference on participants’ behaviors and performance in each design method and led us to suggest how to conduct them in East Asian culture. Mobile Observation Analyzer and Wi-Pro, user research tools we invented to capture user behaviors and needs especially in their mobile context, were also introduced.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Cultural Design Beyond Aesthetic Stereotypes

The late Professor Jay Doblin (1920–1988) stated “Product is frozen information.” The phrase tells us that what is more important is not product itself – while it generally is the primary goal of designers – but the information reflected into product. The information in regards to cultural value of society, life, and user manifests into product. However, designers often merely focus on product itself; form, color, material, or any other physical attribute. Naturally when we – designers – talk about cultural design, we refer to the aesthetic stereotype such as nationwide preferred shape or color. “Rigor” for German, “flair” for Italian, “compact and cute” for Japanese, and “tail fins” for American are good examples for such ideal preferences. Whereas aesthetic differences and similarities among cultures are obviously one of the very important issues in cultural design, the issue of culture in design should be extended to the invisible “frozen information” beyond material and phenomenal culture. As an example, I would like to share a story of my own which goes back to the time I lived in Chicago, USA I bought one piece of frozen pizza and the recipe on the top of its container said “To enjoy the best taste, preheat the oven and put the pizza on the second rack.” I took for granted that “the second” meant “from the top” but very interestingly, or rather surprisingly, most of my American colleagues counted it “from the bottom”! This startling difference of conceptual model puts significant influences on the way that users interact with products. Particularly nowadays, the cultural affordance plays critical role as interaction and interface design have become one of the most important factors in digital information appliances, mobile devices, and web sites than ever before. However, the invisibility of user’s conceptual model is not easy to deal with since designers are mainly concerned with visible problems.

When addressing the problem of invisibility in culture, two issues are raised about design: the users’ cognitive styles in different cultures and design methods for culture. Firstly, the users’ cognitive styles are strongly associated with ways of approaching a task. For the same task, user with a different cognitive style may have a different perception on how to interact with the product. Many researchers, including Nisbett, have proved through various experiments that Asians and Westerners have different cognitive styles; Asians tend to think thematically (relationally) whereas Westerners tend to think functionally (individually). Secondly, a very critical topic in culture and design is about design methods for cultural design. Typically, most major design methods for cultural design are limited in the designer’s personal intuition or, at best, surveys or interviews. These conventional methods cannot effectively work for identifying cultural characteristics or for applying them to design because of the invisibility of culture. The other issue related with design methods for culture is about “different methods for different culture.” For understanding a particular culture, the user is one of the most significant elements. However, design methods developed for a specific culture and in which only the native users have been taken into consideration may not effectively work in a different culture simply because the users are different. This paper introduces and demonstrates several cases of researches regarding these two subjects of cognitive style and design methods for culture carried out in HCIDL (Human Centered Interaction Design Lab), KAIST, Korea.

The Structure of Culture

The myopic definition of culture comes from the wide diversity of definitions of culture. A brief look-up of the word “culture” at Google generates as many as 49,000 results. Images shown are mostly “high arts,” “fashion,” “pop,” or “tradition” implying “a social class” or “classy taste.” We frequently hear someone say “He is a pop culture celebrity.” However, definitions of culture which are widely used throughout the academic society are related to the “behavioral” and “cognitive” aspects of culture; “Culture consists in patterned ways of thinking, feeling, and reacting, acquired and transmitted mainly by symbols, constituting the distinctive achievements of human groups, including their embodiments in artifacts.” This definition leads to the model of culture having the different depths or levels, which many culture-related researchers agreed on. Table 1 shows different models of cultural structure by various researchers (Lee, 2001).

Table 1 Comparison of models of cultural structures

Table 1 can be summarized by the diagram of Fig. 1 having three levels of “Artifact,” “Value,” and “Basic Assumptions.” As shown in Fig. 1, three levels differ from each other in the degree of observability, concreteness, and consciousness. Top layer of culture, “Artifact” consists of the observable, objective, concrete, and tangible elements of culture such as language, food, housing, monuments, clothing, tools, arts and all artifacts human-created. People can consciously and explicitly describe this phenomenal layer of culture. This layer of cultural element – Artifact – is the manifested symbol of deeper layers of cultural element which reminds us of the phrase, “Product is frozen information.” Middle layer of cultural element “Value” includes something that people know but cannot exclusively talk about or elaborate. People know clearly what they prefer or not but cannot articulate the reason. Finally, as we come down to the bottom line, the deepest layer of cultural element “Basic Assumptions” consists of things in people’s minds that are out of conscious awareness, taken for granted, and difficult to know or elaborate.

Fig. 1

Cycling and mutually complementary relationship between different levels of cultural elements

These three levels of cultural elements are mutually reinforcing through the intertwined cycle. The top level of artifacts is linked to the bottom level of basic assumptions through the middle layer of value and norm. Artifacts are embedded in norm and value which are in turn embedded in basic assumptions. The artifacts impart meaning to behavior through the rules prescribed by norm and value, but the meaning of behavior only makes sense in the context of basic assumptions that surround such behavior. This cyclic model also reveals the process of how the product is evolved. At first, a new product is introduced to the real world as a functional artifact (e.g., car for transporting), and later people get to form their individual values of its image (e.g., “sexy” car). Next, if the value lasts long enough to be shared by a society, it would be gradually absorbed deep into a subconscious level (e.g., car as society’s icon). Once the cycle reaches the bottom level, we really need creative people to cross the chasm and start the reverse cycle, i.e., from basic assumption (frozen information) to artifact (product) level. Why “creative” people? Because the core of creativity is nothing but breaking fixed mental blocks (basic assumptions) everybody takes for granted. So far, the first half of the cycle has been dealt with by social scientists, typically in the field of marketing. They go out to the market (artifact or street), carry out marketing research on people, and, consequently, results made of very abstract keywords come out, like, “Trend of back to the basic.” In other words, “information” melts out of a “product.” Then the result of market research (the melted information) is handed over to designers who hardly share the real meaning and background of the information from the former half of the cycle. Naturally, designers handed over with tons of marketing data begin to “freeze the information into product” without understanding the true message of that marketing data. In the future, designers will be increasingly required to be capable of connecting these two half-cycles by crossing over the chasm between them.

Cultural Cognitive Styles and User Interface Design

Cultural Cognitive Styles

Anthropological and psychological studies continue to suggest that cognitive style is culturally different. Particularly, Nisbett reports plausible evidence of such cultural difference, empirically supporting what has been extensively asserted in other disciplines by explaining where this cognitive difference comes from. Nisbett and his colleagues have found cognitive differences between East Asians and Westerners in regard to perception, attention, categorization, and inference.

Masuda and Nisbett (2001) found perceptual differences between East Asians and Westerners through an experiment where they presented animated vignettes of underwater scenes to Japanese and American participants, and asked the participants to report what they had seen from memory. The result indicated Japanese participants to be more likely to report field information such as the color of the water, plant formation, and inert animals than Americans. This study revealed that East Asians are more focused on the context and the relationships among objects, whereas Westerners are more focused on central objects and tend to detach objects from the context.

Concerning ways of organizing the world, East Asians tend to group objects on the basis of similarities and on the relationship among the objects, whereas Westerners tend to group the objects on the basis of categories and rules. Chiu (1972) showed three images of objects (a cow, a chicken, and grass) to American and Chinese children and asked them to make two groups. Chiu found that American children grouped a chicken and a cow together because they are both animals while Chinese grouped a cow and grass together because a cow eats grass (Fig. 2). In the experiment judging similarity that Norenzayan and his colleagues (2002) conducted, Korean students tended to associate a target image with Group A which shares family resemblance, whereas American students appeared to choose Group B based on the consistent property of a straight stem (Fig. 3).

Fig. 2

Example of stimuli for categorization experiment (Nisbett 2003)

Fig. 3

Example of stimuli for similarity judgment experiment (Nisbett 2003)

A central idea to this research is summarized as “Holistic versus Analytic thought.” In their paper “Culture and Cognition” Nisbett and Norenzayan (2002) proposed that cognitive processes differ according to Holistic and Analytic perspectives. They stated that cultural differences in cognitive processes are tied to cultural differences in basic assumptions about the nature of the world (i.e., Holistic vs. Analytic). Holistic and Analytic reasoning was summarized as follows.

Holistic thought involves: (1) orientation to the context or field as a whole, including attention to the relationships between a focal object and the field; (2) a preference for explaining/predicting events on the basis of such relationships; (3) an approach that relies on an experience-based knowledge rather than abstract logic and the dialectical; 4) an emphasis on change, recognition of contradiction, and the need for multiple perspectives.

Analytic thought includes: (1) a detachment of the object from its context; (2) a tendency to focus on the attributes of the object in order to assign it to categories; (3) a preference for using rules about the categories to explain and predict the object’s behavior; (4) inferences that rest in part on the de-contextualization of structure from content, use of formal logic, and avoidance of contradiction.

Cultural Cognitive Styles and Mobile Phone UI

Based on the theoretical background of culturally influenced cognitive differences mentioned in the previous section, which seem more closely related to the matter of communication known as “Human-Computer Interaction,” Kim (2007) in her study aimed to illustrate how culturally different cognitive styles influence the information structure in the mobile phone interface by examining user performance and attitude toward the interface. She hypothesized that the user’s performances and favorable attitudes would be enhanced when a menu structure works or looks like their cognitive style. More precisely the assumption was that Holistic thinkers would take and prefer a thematic menu approach, whereas Analytic thinkers would take and prefer a functional menu approach. For instance, to perform a task of setting a ring tone, the user may go through the “setting” menu in the main screen and find an option for ring tone. Alternatively, the user may find the same option under the “sound” menu in the main screen. Having the option under the “setting” menu is a menu structure, which is “functionally” grouped by the common function “setting”. Having the same option under the “sound” menu is a menu structure, which is “thematically” grouped by the shared context (or theme) “sound.” Considering that Holistic and Analytic thinkers categorize things differently, it was predicted that they would show different behaviors and attitudes toward certain menu structures due to their cognitive difference.

To verify this hypothesis, a prototype test and a cognitive style test were conducted with 30 Korean users from KAIST and 30 Dutch participants from TU Eindhoven. It was considered that their cognitive styles may not be reconciled with their culturally different cognitive styles, as was assumed. In other words, individual cognitive styles may be more salient compared to collective cognitive styles due to the small sample size. For this reason, each participant was asked to perform the cognitive style test so that the correlation between individual cognitive styles and individual performances could be revealed.

In the mobile phone prototype (an interactive prototype running in a desktop environment), the main screen of the mobile phone consisted of six menus (Call history, Messaging, Phonebook, Sound, Display, and Settings). The phone’s setup was possible to change through thematic menus such as “Sound” and “Display” as well as through the “Setting” menu. For example, to set or change the wallpaper on the mobile phone, participants could start from the “Display” menu to select a certain picture from “My pictures” and then click “Set as wallpaper” from among the options in a context menu that popped up from the right side at the bottom. Participants were also able to change the wallpaper by accessing the “Setting” menu, entering “Wallpaper” in “Display,” and finally selecting one picture from among a list of pictures (Fig. 4).

Fig. 4

Two approaches for setting the wallpaper (Top: from ‘Display’, Bottom: from ‘Settings’)

Participants were asked to change the ring tone (Task 1) and the wallpaper (Task 2) to a specific one for each task. After completing the two tasks, they were asked to perform the same tasks again using the other way. For example, if a participant changed the ring tone using the “Sound” menu during the first trial, the same participant would need to do the same task using the “Setting” menu within the second trial. This was to allow the subjects to experience and compare two different ways (thematic approach vs. functional approach) so that they were enabled to choose a preferred approach at the end of the test.

The cognitive style test was intended to discover whether an individual’s categorization style is taxonomic or relational. One target picture and two alternative pictures were presented together and participants were asked to select the one alternative that best matches the given target picture as quickly as possible (Fig. 5). The two alternatives were one that belonged to the same taxonomy as the target picture and one that shared a relationship with the target picture. For example, a picture of a mouse (or a rat) was presented as a target picture, and a squirrel and a piece of cheese were presented as alternatives. A mouse and a squirrel are both animals and so were considered to be in the same taxonomic category. A mouse eats cheese, thus the mouse and cheese were considered to share a relationship. The pictures remained on the screen until participants made their choices. Twenty six sets were presented in sequence.

Fig. 5

Cognitive style test (sample screen)

Cognitive styles were found to be different according to the selected menus. Their categorization tendency was different between a group selecting the “Setting” menu and a group selecting the “Sound” menu while changing the ring tone. Such tendentious differences were also found for the changing of the wallpaper between a group selecting “Setting” menu and a group selecting “Display” menu. Categorization tendency also appeared to be different according to the preferred menus. In other words, the group that selected/preferred the “Setting” menu had a tendency to be more taxonomic (Analytic) than the group that selected/preferred the “Sound” or “Display” menu in both tasks.

Korean participants preferred a thematically grouped menu and Dutch participants preferred a functionally grouped menu. The categorization tendency of the Korean group was found to be more relational compared to the Dutch group, but the tendentious difference was not statistically significant. The sample size was not large enough to be generalized down to a collective cognitive style; for this reason, individual cognitive styles appear more highlighted than collective cognitive styles. However, a number of cultural psychological studies have showed such collective cognitive styles. Thus, the correlation between individual cognitive style and menu structure found in this research can feasibly be applied to a cultural level under the assumption that East Asians tend to make more relational groupings compared to Westerners.

This study shows the possibility of applying a culturally different cognition model into the interface architecture. Current structure of the interface appears universal across cultural areas. However, the findings in this research may be helpful to design an interface suitable to each cultural area based on the fact that the cognitive styles of East Asians and Westerners differ from one another. Given that mobile devices have a limited number of menus due to their small screen size, it is necessary to organize the limited number of main menus appropriately in order to offer logical and quick access to any command or option. This study suggests that menus can be organized in a different way depending on users’ cognitive styles; hence, suggestions regarding structuring items in mobile phone are made available by this study.

The participants having a relational-holistic cognitive style performed the tasks using a thematic approach and preferred this approach in a situation where both approaches were available. Therefore, for East Asian users who are known to be more relational, it may be better to organize menus thematically so that they make more natural and efficient use of their devices. In the thematic structure, options are usually shown in the form of a context menu to offer contextual accessibility, an approach that suits the attributes of a holistic thinker. For example, it is possible to organize main menus with thematically grouped menus such as “Sound” and “Display” and put all tasks related to the theme together.

Participants with a taxonomic-analytic cognitive style performed the tasks using a functional approach and preferred this approach in a situation where both approaches were available. Therefore, for Western users who are known to be more goal-oriented, it may be better to organize menus according to goals or functions so that they feel certain of goal achievement. For example, it is possible to organize main menus with functionally grouped menus such as “Setting” and “Download.”

Cultural Cognitive Styles and Web page UI

It is also important to consider the impact of culturally different cognitive styles on web page user interface. Web sites-the most popular medium of global communication-can potentially be visited by people around the world and people from different cultures employ different usage strategies on the web site. Cognitive style plays an important role in accommodating an individual’s typical mode of perception, thinking, remembering, and problem solving in order to promote web usability.

Dong and Lee (2008) proposed a new approach to enhance the usability of web page by applying the culturally different cognitive styles in their study. They expected East Asians and Westerners to apply different viewing patterns and perception styles while browsing web page. To prove this assumption they conducted an eye-tracking experiment to observe and compare viewing styles of different cultural groups while viewing web page.

The prototype used in the experiment was designed by imitating a popular web site, Yahoo! The web page prototype was designed with the most basic web page elements and page layout. The clearly and neatly divided areas were designed to easily allocate eye movements data. Stylization of the design was restrained to limit distraction to the participants (Fig. 6). Prototypes with identical contents and layout, as well as identical page elements, were designed. Three different language versions of the prototype were provided in English, Chinese, and Korean.

Fig. 6

English(left), Chinese(middle) and Korean(right) version of the prototype

14 American, 15 Chinese, and 12 Korean participants were invited to take part in the experiment. Participants were exposed to the prototype version in their native language, and they were asked to freely look at the web page without clicking on anything since the task was trying to let people show how they actually view a web page without a specific searching item so that their natural viewing pattern could be revealed. As soon as the prototype was shown on the display, the eye-tracking device was triggered by the experimenter to record the eye movements, and the recording was stopped after 30 s.

Each eye-tracking map from the three different national groups (Fig. 7) was marked in a chart according to two criteria of analysis. Criterion 1 was to analyze the participant’s general reading style, whereas criterion 2 was to reveal the exact pattern behind that style. Observed viewing styles and patterns were codified and mapped for each national group in two radar maps (Figs. 8 and 9).

Fig. 7

English (left), Chinese (middle), and Korean (right) eye-tracking samples

Fig. 8

Mapping viewing patterns according to analysis criterion 1

Fig. 9

Mapping viewing patterns according to analysis criterion 2

Figure 8 indicates that each group has a moderately different viewing pattern. For example, 7 out of 9 American participants tended to read the prototype page in sequential order, while only few Chinese and Korean participants showed sequential reading patterns. On the other hand, Chinese and Korean participants were more likely to scan back and forth between page contents, and they were more likely to scan the page in a circular pattern. A web page can be perceived as a package of different informative objects such as information items and/or information boxes, arranged on a field. Holistically minded people have the tendency to see the field as a whole. Hence, they employ a strategy to perceive the web page by scanning across all information objects. Scanning back and forth implies that Chinese and Koreans are not really reading carefully, but just randomly scanning the page. Analytically minded people, on the other hand, tend to detach objects from their background field. Those people tend to focus on each piece of information one by one; a behavior, which leads to a sequential reading pattern. Americans seldom scan without examining the details and rarely scan back and forth between contents. Americans are likely to focus on the page title and also likely to read the navigation, while few Chinese and Koreans do so. Analytically minded people are inclined to think in categories, so knowing what kind of categories the web site has would help them to perceive the web site.

Figure 9 illustrates that most Chinese and Korean participants applied a “0” shaped viewing pattern, while Americans rather applied a “5” shaped eye movement on the page. “0” shape implies that Chinese and Koreans tend to scan the whole page which is similar to the circular scan above. Most Americans show a tendency to read from the center to the periphery of the page. Other viewing patterns in Analysis Criteria 2 do not seem to be significantly employed by a certain group.

The present findings indicate that Holistically minded people and Analytically minded people have unique ways of perceiving a web page. The characteristics of perception reflect some aspects of Nisbett’s proposition about cognition. It is suggested that the web page designer should be aware of the cognitive differences existing among Holistically minded people and Analytically minded people. Consequently, web page design must be carried out according to the target audiences’ specific cognitive style in order to enhance perception and usage of the web page. This study further prepossessed several recommendations for web page design as shown in Table 2.

Table 2 Recommendations for web page design for different cultural group

Design Method for Culture

We understand the importance of having design or user experience reflect cultural values and possible differences. However, understanding users properly in their cultural context should come first. In this part, we will introduce some of our studies on localized design methods for specific cultural groups and user research tools we invented to capture user behaviors and needs especially in their mobile context.

Localization of Design Methods for Different Cultures

Cultural Difference in Communication Pattern

Individualism versus collectivism is an idea that contrasts an individual who only cares for oneself and one’s direct family members(I-conscious) to the one who emphasizes the importance of loyalty and unity for the group that cares for one(We-conscious) (Hofstede 1991). This idea is related to the communication pattern of the society’s constituents and it can be explained in relation to Hall’s (1977) “context” theories (Hofstede 2001). In Hall’s culture theory, information during communication or in a message is a part of context. It is more or less defined by the degree to which the message or communication is internalized by an individual. In “high context cultures” most information is included in the context, and thus expressed less externally. In contrast, communication is direct, clear, and expressed externally in “low context cultures.” Hofstede (2001) revealed that high-context communication occurs in collectivistic culture and low-context communication occurs in individualistic culture.

Some cross-cultural studies in cognitive psychology and creativity research explain that individualistic and collectivistic cultures have different attitudes toward discussion or argumentation because of their values and education systems. Richard Nisbett (2003) says that “lively discussion” is a part of the culture in individualistic countries, supporting academic activities and formulating social systems. People in individualistic countries learn to argue and persuade from childhood on and believe that problems can be solved through discussion.

Differences in Focus Group Interviews in East Asia and Europe

The degree of how much people care about a “positive face – the desire to be liked and approved by other people” (Ting-Toomey 1998) – can influence the degree of participation in different design methods where interaction between different individuals is required. In focus group interviews, participants are asked to talk about their personal experiences and subjective opinions on certain topics in a group. When people care about their positive face, they may be afraid that their experiences or opinions might be “incorrect” or sound silly. These concerns could result in passive participation styles. Moreover, respective participants might be reluctant to present negative opinions and will have the tendency to preserve others’ face, too. In consequence they might hold back disapprovals of any kind even though problems with products under evaluation were encountered or disagreements with other participants exist.

Chavan (2005) argued that people from collectivistic cultures tend to work around problems rather than to criticize them, while people from individualistic culture are likely to find problems and criticize. She claims that this difference may “corrupt” interview data when not carefully considered. In their comparative experiments on usability tests, Hall et al. (2004) revealed that a think-aloud protocol did not work well with people from collectivistic cultures, because the subjects felt uncomfortable to point out problems.

Lee and Lee (2007) explored cultural difference on participants’ behaviors in focus group interviews and suggest how to conduct them in collectivistic culture. They hypothesized that people from collectivistic/high-context cultures will show less activeness of participation and weaker interaction among members than participants from individual/low-context cultures in group interviews, and conducted comparative experiments with participants from two cultures, the Netherlands and Korea, employing “focus group interviews.”

Five engineering students in their early twenties (three males and two females in the Netherlands and two males and three females in Korea) were recruited in each country. In two focus group sessions in each country, the facilitator took a minimized role limited to giving topics and distributing turns in order to allow group-centered interaction and to minimize her influences on participants’ interaction patterns. To elicit different kinds of speech styles such as storytelling or argumentation, the questions for the interviews included speaking of their own experiences of digital multimedia use, discussion of the products that were previously tested in individual usability tests, and the participants’ desires for future products. Each session lasted about 110 min including a 10-min break. Each session was video-recorded for further analysis.

The analysis mainly aimed at comparing the degree of participation and interaction patterns among participants in two countries. To support comparison, Lee visualized the amount of participants’ speech, besides describing observed findings. At first, she divided the interview time into 30 s interval units and highlighted those units every time each participant spoke (see Figs. 10 and 11). To see how they argue with each other, three categories were set into “questioning,” “approval,” and “disapproval” and the corresponding utterances were represented as three different types of arrows. The arrows start from a person reacting and head to a pointed person. Facilitator’s utterances were also coded into four categories, such as “providing a topic,” “pointing out a person,” “asking volunteering,” and “detail questioning,” to map what kinds of facilitator roles are required according to participants’ activeness and interaction patterns.

Fig. 10

Timeline analysis of the group interview in the Netherlands

Fig. 11

Timeline analysis of the group interview in Korea

In Figs. 10 and 11, the number of highlighted areas represent the amount of speech and the number of arrows the degree of member-to-member interaction. Lee also expected to observe a change in the degree of participation and interaction over time. To easily compare the amount of speech and interaction as well as the difference of individual participation in a group, she calculated the number of highlighted time units and presented them as circular areas. Thus, the radius of the circle as shown in Fig. 12 is defined by the amount of highlighted units of Figs. 10 or 11, respectively. The number of member-to-member interactions was also presented as the width of lines according to each category.

Fig. 12

Diagrams of the amount of speech and interaction patterns (a) in the Netherlands and (b) in Korea

Firstly, more speech was observed from Dutch participants than Korean participants, even though Dutch participants spoke in their second language. When the facilitator provided topics Dutch participants told “stories” and “anecdotes” related to the topics, while Korean participants tended to make short “answers” except for participant “K2”. This tendency could already be observed when participants introduced themselves in the beginning of each session. For example, Dutch participants talked about their current studies or hobbies and, by doing that, broke the ice by themselves, while Korean participants told only their names and majors. When the facilitator asked what kind of digital devices participants possess, one Dutch participant told stories of his mobile phone, such as when he bought it, what he likes and dislikes about it and even the subscription he had. Then another participant responded to it by telling his story about getting his phone from his brother and the moment when he almost broke it. Therefore, the facilitator was not engaged to ask too many detailed questions. Moreover, she sometimes needed to skip some prepared topics since Dutch participants already talked about stories related to those topics.

However, in Korea participants answered relatively in short, for example, “I have a mobile phone and an electronic dictionary. I do not use an MP3 player.” Then, the facilitator needed to ask “why” to know the reason for not using an MP3 player, and the Korean participant answered, “because I gave it to my mom.” In effect, the facilitator was required to provide topics and ask detailed questions in Korea in order to elicit the information needed from participants.

Interactions among participants were observed to be considerably lower in Korea as shown in Figs. 10 and 11. On the other hand, Dutch participants showed lively discussion and member-to-member interaction during the whole session. They actively added opinions, asked questions, and showed approval and disapproval to other participants’ utterances. It was observed that the turn-taking speed was faster in the Netherlands and, in fact, Dutch participants often interfered with each other, while this kind of interference seldom happened in Korea. In the Netherlands, simultaneous subgroup discussions sometimes happened naturally due to intensive argumentation. In contrast with Dutch participants, Korean participants seemed to “answer” the facilitator’s questions rather than to ‘‘discuss” with each other. It was also observed that Korean participants tended to look at the facilitator while speaking while Dutch participants had many eye contacts with other participants. Even though the Koreans’ interaction was passive at the beginning, it increased as time went by (Fig. 11). The degree of participation of each member in a group was less equal in Korea (Fig. 12).

The results of the comparative experiment show different behaviors in focus group interviews between two cultures. More active participation and member-to-member interaction was observed in the Netherlands, as hypothesized based on literature studies. Korean participants heavily relied on the facilitator while Dutch participants started active discussions voluntarily. Member-to-member interactions in Korea increased in the latter part of the interview. It is assumed that the participants became more accustomed with the discussion as time went by. If Korean participants need more time to get comfortable with the interview situation and other participants, ice breaking will play an especially important role for them.

The findings revealed that “instigation” and “motivation” for speech from participants and member-to-member interaction are more required in focus group interviews in collectivistic culture. This “instigation” and “motivation” can be facilitated by the moderator’s skills, provided scripts and activities, physical environments, and so on. To explore how to facilitate “instigation” and “motivation” in focus group interviews in collectivistic culture, various tools can be applied in a session, such as pre-question cards, Mini-me dolls, bottle spinning game, and TV home shopping setting. Lee applied several of those tools in additional research and found that these tools provided “stimuli” which can boost participants’ interests and motivation. Participants from collectivistic culture easily reflected themselves into an imaginary setting, for example, TV home shopping situation. This kind of role play and imaginary setting can play an important role to elicit active discussions, as it enables participants to adopt a rather indirect communication style.

Based on findings gathered from the empirical cases, Lee suggested some promising factors worth further development for deriving collective knowledge from collectivistic culture:.

• Foster sensitivity and motivation by providing playful props and activities.

Utilizing playful stimuli can allow participants from collectivistic culture to feel comfortable with the interview situation and provide them with motivation.

• Facilitate indirectness by providing imaginary roles and settings.

Participants from collective culture tend to adopt indirectness of their communication styles and get empowered in role playing and imaginary situations.

• Ice breaking is especially important for collective culture.

Participants from collectivistic culture seem to need more time to get accustomed to the interview situation and to interact with others. Try to open dialogues before the interview by pre-tasks or informal meetings. Playful props and activities will also help to break the ice in the beginning.

• Place tasks of evaluation and critiques in the latter part of the interview.

Participants from collectivistic culture tend to be reserved in the early stage of the interview. However, they become more active once they gain familiarity with the interview situation and other participants. Place tasks requiring criticism in the latter stage.

• Visualize respect for their participation and information.

To show approval and respect of participants’ opinions will give them confidence and motivation.

Cultural Effects on Idea Generative Session

Generative session is a user research method that leads people to express their experiences by making a collage with given images; an approach particularly suitable to communicate latent knowledge which cannot easily be expressed explicitly. In generative sessions, images provided for the collage stimulate people’s latent thoughts and needs and help to describe them. Therefore, the application of stimuli images plays a very important role for a successful creative workshop.

You (2009)) explored how to enhance the performance of generative sessions considering the fact that East Asians have the tendency to focus on the relationships of objects to context and to see the field as a whole, opposed to Westerners who tend to focus on the objects and detach an object from its context. Based on the cognitive difference in perception, she could make a hypothesis that cultural cognitive style – holistic (Eastern culture) and analytic (Western culture) – affects the way of using stimuli images and the result of generative sessions.

For the collage experiment, 125 stimuli images with central object and background were selected from magazines. Two sets of stimuli images were prepared – one with central object and background, and the other one with only central object and white background (Fig. 13). In total, 36 people were invited consisting of 19 Koreans and 17 Europeans. Participants were asked to describe their “photo-taking” experience by making a collage with one of the two stimuli image sets (one with central object and background, or the other with only central object), then presented his or her creation to the experimenter. The task was open-ended, and took around 1 h.

Fig. 13

An example of stimuli images with background (Top), and with only central object (Bottom)

After the collage experiment, participants rated “how happy the central figure on each cartoon looks” on a 10-point scale in the facial emotion judgment test (Masuda 2008) which was to verify the participant’s cultural cognitive style (Fig. 14). Of the 56 original cartoon sets of Masuda’s experiment, 16 images were used: (a) two different central figures (Asian and Caucasian), (b) with two different expressions (moderate and intense happiness), (c) and peripheral figures with one of four different expressions (anger, sadness, happiness, and neutral). Each cartoon was presented in random order on a computer monitor.

Fig. 14

Happiness cartoon set (Masuda 2008)

The discrepancy of the Korean participants’ judgment of matched images (a happy target and happy backgrounds) and that of mismatched images (e.g., a happy target and angry backgrounds) was significantly larger than with European participants. These results indicated that the Koreans were more influenced by the changes in background facial expressions than the Europeans. It also appeared that the Koreans and the Europeans have significantly different cognitive styles – Koreans were more likely holistic, whereas Europeans were analytic.

The performance of collage was evaluated by the richness of contextual information in a whole description. Content analysis was used to examine the participants’ description. Table 3 shows topics extracted from collage and presentation. The frequency of each topic from each participant group was compared and significant difference was found especially in the frequency of describing “what to take – object.”

Table 3 Topics from collages and presentations

The Korean participants were more affected by background than the Europeans and were induced to talk about contexts. It means that background of images was a means for the Koreans to describe contexts. Therefore, when context is important for research, background of images can be used to stimulate the participants’ contextual memories. In addition, background contents should be diverse so as to extract people’s various contexts. However, when context is not important for research, background on images should be minimized so as not to distract people’s attention for a task. On the other hand, the European participants tended not to be affected by background and they explained mainly what to take and fewer contexts than the Koreans. Thus, when context is important for research, more instruction for describing contexts or stimuli words like “when,” “where,” or “why” are needed to extract people’s contextual memories. Image contents should be more varied than backgrounds.

The results also showed that the background of images stimulated more thoughts and feelings from the Koreans than from the Europeans. Therefore, when thoughts and feelings of Korean participants are important for research, background of images can be used to stimulate participants’ thoughts and feelings. In addition, background content should be diverse so as to extract people’s various thoughts and feelings. However, when thoughts and feelings of people are not important for research, background on images should be minimized to not distract the participants’ attention for the task. On the other hand, European participants tended not to be affected by background. Thus, when thoughts and feelings of people are important for research, more instructions for describing thoughts and feelings or stimuli words like “I think that,” “cute,” or “conventional” are needed to extract people’s thoughts and feelings.

Both the Korean and European sample mentioned “photo style-how to take” a lot. This showed that photo-style of stimuli images themselves stimulated people to talk about photo-style. Thus, to gain a certain sort of information from people, appropriate stimuli to describe the information are needed. In addition, not just image content but also the style can stimulate people’s thought.

Tools for Understanding the Invisibility of Culture

Another important issue of design methods related with culture and design is how to understand the invisible part of culture. Latent and tacit needs cannot be explicitly articulated because users take it for granted and are not even aware of it. The user observation is one of the most frequently used design methods for overcoming these difficulties. However, users using mobile devices are extremely difficult because users are continually moving and interaction with mobile device is too micro to be observed. Two tools were developed to facilitate the observation of users using mobile devices in HCIDL. MOA (Mobile Observation Analyzer) and Wi-Pro are introduced in the following section.

MOA (Mobile Observation Analyzer)

MOA was developed for understanding the user’s behavior in a real-usage situation in order to identify tacit and latent needs. The key factors for success in this stage rely heavily on two issues: how to conduct user research as naturally as possible so that the user reveals his tacit or latent needs in an uninterrupted environment; how to understand users at different levels of activity so that a researcher has a systematic understanding of users without missing any critical aspect of their needs. MOA allows researchers to observe users from three points of views for exhaustive and comprehensive understanding of users: second point of view, first point of view, and third point of view (Fig. 15). First, MOA adopts the technology of a micro-wearable camera to understand the user’s self point of view, what the user sees and interacts with at a very micro level. A very small video camera developed as a medical endoscope is embedded into the user’s eyeglass frame. Wearing the eyeglasses as usual, without attracting other people’s attention, the designer can get the video data of what the users see and can observe and record their interactions with mobile devices automatically and naturally (Fig. 16). Secondly, the shadow-tracking method is also implemented to understand a second point of view; here the researcher observes the user in close proximity, seeing the user’s gesture of holding and carrying around the mobile device. The shadow-tracking method is a very typical observation technique. Finally, GPS technology is used for a third point of view, to observe the user’s macro behavior, like the users’ movement, rhythms, and position as they move about town.

Fig. 15

MOA system structure

Fig. 16

Observation system with wearable camera

MOA also has computer-aided video-annotating software as shown in Fig. 17 which allows a researcher to simultaneously view video clips of all three points of view. It allows a researcher to have a comprehensive understanding of the user’s behavior from micro to macro level; understanding what the user saw, what they clicked on, during what motion sequence, and how they moved around. Its interface consists of four components: “tool bar,” “video controller,” “event table,” and “editor.” The tool bar manages all the software by providing access to basic menus like “Open,” “Save,” “Play,” “Stamp,” “Set Focus Window,” and “Close.” The video controller opens the video clips recorded in three different views of self, first, and second, and controls “Play,” “Pause,” and “Stop” like on a VCR. While playing video clips, designers analyze various users’ behaviors and use the “Stamp’’ function which allows him to stop and to record specific parts of the video clip for annotating user behavior in a predefined framework. All the stamped portions of video clips are to be recorded as “events” in “Event Table” in sequential order. In the Event Table all the events stamped in three video controllers are listed up in sequential order with time, captured still image, analysis frameworks like “user,” “object,” “action,” etc. through which a designer can annotate the stamped user behavior. Designers can search for specific events by simply clicking on the relevant images. The editor allows designers to have an extensive analysis and a description of a particular event he or she selected. Through the analyzer and with all these elements mentioned thus far, a designer can effectively annotate and sort different data to capture insightful understandings of user behavior in a mobile environment. An example of the full user behavior scenario drawn from MOA is shown in Fig. 18.

Fig. 17

MOA video annotating software

Fig. 18

Total scenario from user behavior observed by MOA

Wi-Pro (Wish Prototyper)

People carry around and use mobile phones anywhere and anytime, and mobile phones become an inseparable part of the user’s paraphernalia, like eye glasses. Obsessive behavior in using the mobile phone has made it unimaginable for the user to have a life without a mobile phone. People take it for granted as a necessity of life. A simple, low-fidelity tool called Wi-Pro was developed to capture a list of wishes of someone without a mobile phone. The user’s mobile phone is taken away for one or two days and instead a fake one that is similar to a real mobile phone in shape with a folding cover is given, rather like a fake cigarette for people trying to quit smoking. People are so obsessed with smoking that they habitually put the fake cigarette to their lips even if it does not deliver nicotine. Similarly, users habitually take the substitute mobile phone out of their pocket even without any actual functions. A binding of Post-It notes inside the dummy mobile phone allows the user to write down their wishes – what they would want to do at that specific time if the mobile phone was real (Fig. 19).

Fig. 19

A dummy mobile phone given to user for Wi-Pro

While the user is playing with the dummy mobile phone, the user’s actual mobile phone is analyzed with the user’s permission to understand his use of the phone, for example, “how does the user personalize the menu” or “how does he organize addresses and other data.”

After 1 or 2 days of Wi-Pro, the user returns the fake mobile phone he used and on which he recorded his wishes. Upon retrieving the fake mobile phone, all the Post-its are arranged in a time sequence as shown in Fig. 20. All the wishes are analyzed in various ways, like frequencies of specific wishes, patterns over time, wishes in specific places, and so on. The user is invited to have a debriefing session and an in-depth interview about the wishes he expressed. This simple, low-fidelity method is very quick and effective in understanding the user’s intense wishes and new ideas.

Fig. 20

An example of timeline analysis of user’s wish notes from Wi-Pro

Conclusion

Due to rapid globalization culture becomes a very critical issue in design. Whereas visible elements like color and shape become more and more universal, invisible elements like people’s conceptual models of product usage are still kept local to their culture. Moreover, the invisible part of a product such as user interface design and interaction design are becoming more important ever since product has become a knowledge-supporting tool. This tendency of nonvisualization is expected to be more accelerating as the way of interacting with product is resuming the more natural way of gesture like Wii from Nintendo. The people’s natural behavior and gesture is one of the most culturally dependent elements and it will require a new dimension of study in cultural design. The more the invisible part of cultural design becomes important the more the localization of design methods is strongly required. Since invisible design problems can be dealt also only with invisible methods like observation and ethnography, design methods developed for observing people in the specific culture need to be customized in different cultures. This paper attempted to study design methods mainly in relation with Nisbett’s theory but the study needs to be further expanded with other cultural dimensions like Hofstede’s or Edward T. Hall’s.