User modeling and profiling has been used to evaluate systems and predict user behaviors for a considerable time. Models and profiles are generally constructed based on studies of users’ behavior patterns, cognitive characteristics, or demographic data and provide an efficient way to present users’ preferences and interests. However, such modeling focuses on users’ interactions with a system and cannot support complicated social interaction, which is the emerging focus of serious games, educational hypermedia systems, experience, and service design. On the other hand, personas are used to portray and represent different groups and types of users and help designers propose suitable solutions in iterative design processes. However, clear guidelines and research approaches for developing useful personas for large-scale and complex social networks have not been well established. In this research, we reflect on three different design studies related to social interaction, experience, and cross-platform service design to discuss multiple ways of identifying both direct users and invisible users in design research. In addition, research methods and attributes to portray users are discussed.
Understanding target users is considered a basic step towards developing good products and services. In traditional industrial design, marketing, city planning, and environmental design, a wide range of research methods, including surveys, field studies, interviews, and focus groups, have been used with the clear purpose of identifying target audiences’ preferences and needs [
To support rapid IT development and iterative design processes, it has become important to have clear images and models to represent end users: ideally models which can be reused and reapplied in the development of different products and services [
With a purpose similar to that of user modeling, personas have also been used to portray user types in serious games, adaptive training systems, communication, and experience design [
Although both user models and personas can offer various benefits in different phases of iterative design, including initial concept development, user testing, and redesign cycles, there are several problems associated with their application in complex designs of social interactions, experiences, and services. First, user-modeling methodology has been developed with a focus on human-computer interaction and centers on task-oriented analysis of behavior patterns, cognitive processes, and demographic data. The approaches and modeling can reveal detailed information about the ways that a user interacts with a system, but are incapable of taking large-scale social contexts into account [
Much work has shown that having social perspectives and understanding nuanced social interaction will be the most challenging but necessary topic in today’s experience and social media design [
This paper discusses current methods, approaches, and frameworks applied in design research, including several user-centered approaches, design frameworks, and models developed for studying cultures. We highlight the strengths and shortcomings of these methods when applied to capture different levels of user information, from detailed interactions to sociocultural backgrounds [
We reflect on the three case studies to recommend suitable user study methods for supporting large-scale services, adaptive systems, and development of serious games, as well as necessary information to gather when portraying users. By understanding participants’ relationships and roles within a broader context, we also derive insights for identifying invisible users and potential users and the importance of understanding different users’ motives and concerns.
Emerging areas of interest in IT development, such as serious games, adaptive systems, and experience design, are concerned with large complex communities and diverse generations. Large-scale services such as e-government and social media inevitably have to support people’s social activities in the coming future. According to the central concept of ethnomethodology, people are intelligent and creative, and with ad hoc practices, they can apply their knowledge across domains and act in different contexts [
In social science, the strategies to study a culture include traditional anthropological approaches, such as participant observations with long-term involvement of the fields and interpretative approaches, in which researchers collect, conceptualize, and induce the concepts through diverse methods. For years, numerous discussions and debates have been held about ways to generate more solid and scientific results while applying these methods and approaches. Many researchers have focused on measurements and standards for improving the validity and reliability of qualitative research [
Applying the above-mentioned positivist or interpretivist approaches to gain knowledge about society and culture is very time consuming and requires experienced researchers to collect and analyze raw data. Therefore, to reduce the cost and to gain insights more efficiently, there are numerous theoretical frameworks and models developed in different domains. For instance, since the 1980s, researchers in management and leadership have discussed numerous models for understanding organizational cultures [
In the fields of management and leadership, many scholars have taken similar approaches to identify cultures—firstly recognizing a culture’s representative characteristics and then categorizing them into types. These researchers generally take an organizational point of view and focus on structures, power distribution, and divisions of labor [
In contrast to categorizing cultures into types, many scholars focus on finding cultural concepts and patterns for intercultural studies. The earliest works are Hall’s books, in which he identified two dimensions of culture, including the high- and low-context communication and polychromic versus monochromic time orientation [
Instead of classifying existent international or organizational cultures into types, some researchers have focused on developing models for analyzing and understanding the culture of a particular group or organization. For instance, Schein [
There is one extreme example of the ambitious objective to combine all aforementioned concepts into one single model. Considering cultural impact on the implementation of enterprise resource planning (ERP), Krumbholz and Maiden [
The models of cultural types provide several ways to identify national or organizational cultures and have indicated the significant relationships between power structure and personal behaviors [
In social interaction and experience design, issues which are related to individual attitudes and values can often be traced back to and better understood in the light of social norms or culture. However, studies of how society and culture act as constraints for design have very different goals than traditional broad studies of culture itself. They also differ from the above-mentioned models and measurements of culture which focus purely on work and whose purposes are either to enhance organizational performance or to improve management and leadership. Due to their strong task-orientation, the methodologies proposed for these studies focus on practical and behavioral views, environments, and symbols. They overlook several cultural aspects which are important in socially motivated interactions between people, technology, and services, such as people’s motives for actions and their emotional needs.
In the past decades, design research of information and system development has focused on interactions between humans and machines. Based on the introduction of cognitive psychology, most studies have concentrated on mental processes and information flows, with the scope being achieving a task or solving a problem. Taking an example of Donald Norman’s seven stages of action, which have had a lasting impact on usability engineering and industrial design, the analysis of human action is concerned with a loop of forming the goal, forming the intention, specifying an action, executing the action, perceiving the state of the world, interpreting the state of the world, and evaluating the outcome [
The scope of design research in technology development expanded during the 1990s from task analysis to more complex activity analysis. For understanding larger-scale usage contexts and supporting user experience design, many ethnographical research methods were introduced in design practices, such as interviews and long-term involvement observations [
Among these research frameworks, contextual design methodology [
Activity theory, with its roots in the 1930s Soviet cultural-historical psychology, was introduced and adapted into HCI and CSCW as a lens in ethnographic research [
With a strong intention to bridge the gap between subjective-to-objective and macro-to-micro concerns in design study, MultiLevel Social Activity Model (MLSAM) was proposed by Huang and Deng [
From an integrated sociological viewpoint, individuals’ behaviors are not only encouraged, but also constrained by norms, religions, and sociocultural backgrounds. To well support hypermedia and adaptive systems for social purposes in the future, design research needs to extend from traditional usability evaluation and task-oriented studies to a larger-scale sociocultural scope. Therefore, an in-depth research approach is needed to answer to the complexity of social interactions. The following section presents three case studies of applying both contextual design [
To better support product and service design in the future, we applied multiple user study methods to uncover the complexity of participants’ perspectives, interactions, and attitudes within different types of activities: a traditional ritual of a Taiwanese tea ceremony, Taiwanese teenagers’ social activities, and technology use, and sports watching in Southern Europe. The first study is selected to show how well-accepted software design methods can be used together with the multilevel social activity model [
The first case is a study of traditional tea ceremonies in Taiwan. These ceremonies have deep cultural roots and also contain complex forms of social interaction, which are typically ignored in most IT development and for which established research methodology is lacking [
In this case study, we applied multiple user experience research methods, which include observations [
The flow model in Figure
Different user types in tea ceremonies.
The flow model of tea ceremonies
The multilevel social activity model of tea ceremonies
While the flow model captures roles and interactions, the multilevel social activity model further reveals how participants’ underlying motives and attitudes toward tea ceremony differ between generations. It also shows that attitudes and resulting behaviors are all strongly rooted in a cultural context. Both the younger generation and the elderly/middle-age generations (both groups in Figure
The multilevel social activity model identifies different user groups by their attitudes towards activities and motives for participation. In addition to people’s direct and immediate requirements, we consider that there are many hidden reasons for people to engage in a social activity that may be overlooked by methodologies that do not go deep enough. For instance, the social activity model reveals that the elderly have positive attitudes toward tea ceremonies potentially reflecting the awareness of Westernization, while the younger generation is more attracted by the health benefits associated with drinking tea.
An understanding of differences in user groups’ cultural backgrounds and underlying motives can help designers make better decisions, in particular in design for social activities and communication. For instance, knowledge of cultural features, for example, pouring a cup of tea to convey esteem and respect during a tea ceremony, can further lead to more accurate portraits of users and better predictions of user behaviors. However, it is difficult to gain this knowledge using only activity- and usage-centered design methods. Instead, the case studies indicate that the knowledge can be gained from a historical and broader-context approach.
Crazy Vote was a social website in Taiwan that provided users with personal web space, such as weblogs and a message board. Due to its unique interface and features for voting on users’ portraits, it became the biggest social website for Taiwanese teenagers in 2008, with more than 20,000 users of ages 15 to 19. To guide future application development, the company supported a two-month research project to fully understand their users’ online activities and expectations of social media [
By extending the research scope with both contextual inquires and in-depth interviews, the study shows that teenagers’ common processes of making friends are complex, but flexible. In addition, the boundary between online social interaction and actual relationships is very blurred to Taiwanese teenaged online users. Their reason for making new friends on the Internet is simply to expand their interpersonal relationships in the real world. For instance, the interviewees mentioned that they preferred to make friends who live nearby to increase chances of meeting up in person, as society does not encourage teenagers to travel alone. Social issues that are associated with meeting online friends also make teenagers form unique networks to ensure that all members are using their real identities and to later develop real-life confidence in each other.
As shown in Figure
Diverse user types of social media.
Some teen users had a common and well-defined procedure for making friends successfully and efficiently in the Crazy Vote platform, and most users in the social-oriented group were aware of and applied this process. First, nonverbal introductions would take place through the voting system or by sending emoticons to others. Communication would then be initiated by leaving a private message or by visiting and leaving public comments on each other’s blogs. People who share similar interests and habits may then exchange other online contact information, such as MSN Messenger or Yahoo! Messenger accounts, and start communicating electronically outside of the Crazy Vote platform. In the end, these online friends may end up talking on mobile phones and meeting up face-to-face.
Rapid expansion of information and communication technology has made young people comfortable with using a wide range of communication platforms. Although taking place in an online environment, the observed process among Taiwanese teenagers for making friends is natural, mature and matches traditional Taiwanese social norms. For instance, the initial use of nonverbal emoticons and “likes” to make others aware of their presence was described in the interviews as a type of “reserved” introduction, similar to a head nod or eye gaze. Young users considered it too aggressive and impolite to suddenly show up and introduce themselves in front of strangers. However, according to both interviews and online tracing, users in the self-oriented group were unaware of this process and fell back on expanding their presentation of themselves in the system. Although both socially and self-oriented users initially shared the common goal of making new friends through the Crazy Vote website, the self-oriented users perceived a difficulty to initiate communication and greet strangers in proper ways, which later caused them to focus on their own blogs.
This study illustrates how more in-depth user research can lead to detailed interaction issues as well as an understanding of sociocultural contexts and their influences on users’ motives and behaviors. Such knowledge can also be applied and reused in many different design projects. In addition, understanding users’ expectations and abilities helps development teams make better decisions and predict user engagement. For instance, the social-oriented group’s capabilities and successful strategies of making friends can be applied in social media design to help and guide the other types of users. However, current design research generally focuses on a single platform or a particular environment, which narrows down the research scope and overlooks people’s great ability to manipulate different resources to achieve their goals.
The third case study investigates people’s sport watching experiences in Portugal. The design process began with interviews and observations of 20 active sports fans and people who had participated regularly in football watching activities, followed by modeling according to MLSAM [
The flow model in Figure
The flow models of sport watching.
The flow model of football watching at home
The flow model of football watching at bars
The multilevel social activity model (Figure
The multilevel social activity model of sport watching.
Using the users types identified in the multilevel social activity model, we marked the primary supporters (football fans and community) in orange in Figures
Due to resource and time limitations in this academic project, only primary supporters were selected for the continued development. On the basis of the research findings, two personas were developed who, while both being primary supporters, have very different expectations of the IT product used. As shown in Figure
The uses of personas in the service design process (red dots).
Having a clear image of two user types also helped designers and engineers generate correct use cases [
In this section, we reflect on the three case studies and argue that two types of user information, the dynamic and static attributes, should be captured and included in design research. The suitable methods and techniques to capture this information will also be discussed.
The cases indicate that there are two types of information that have not been highlighted in general user modeling, profiling, or persona description, yet are important for identifying potential users and predicting their behaviors. The first type of user information is dynamic attributes, such as knowledge space [
The importance of the first type of information emerged from both the case studies of tea ceremonies and of teenagers’ social activities. As shown in Figure
In the case study of teenagers’ social activities, we found that although teenagers are willing to be active users of technology in their pursuit of new friends, their available modes of communication are often limited by their economic status. As teenagers generally have limited financial resources, most IT products and media producers are unwilling to develop services for them. However, the older teenagers will see great improvements in their personal economy in only one or two years, and their great interest and reliance on the Internet will make them the target audience of smart phones, tablets, and other IT products in the near future. This type of information, predicting future audiences, is basic in marketing, but it is often overlooked in design research.
Different from most work in adaptive system development [
More permanent and static information, including individuals’ preferences, attitudes, and values, has been highlighted in some work of developing personas and user models [
According to multilevel social activity models and flow models of all the case studies, participants in a certain activity can have very similar behaviors, but their concerns, attitudes, and motives may still differ greatly. For instance, in the case study of football watching, we identified that participants’ requirements and underlying motives varied, including seeking high-quality watching experiences (primary supporters), seeking detailed information (primary supporters), wanting to discuss the game with others, and enjoying each others’ company (potential supporters). However, customer journey maps [
The activity journey map.
In addition, understanding of high-level information such as values, attitudes, and sociocultural background is easily taken for granted and therefore ignored in design research. Mulwa et al. [
On the basis of the studies of different social activities in different countries, we have discussed the strengths and weaknesses of different research methods. To better support rapid design research in most IT development, common user-centered design methods and frameworks provide a cheaper and more efficient way to account for the behaviors and interests of target users. However, in design for supporting communication, social purpose, and larger-scale services, we recommend to apply multiple research methods to gain deeper insights into the contexts and to identify potential and invisible users. Therefore, we propose the following process. Focus on a certain activity: as a pilot study, researchers can choose a representative activity and apply common context-, user-, or usage-centered research methods to capture events in great detail. In this stage, people who actually participated in the activity, objects, media, environmental circumstances, and interactions are identified. Extend the context: the goal of this stage is to identify the flow of information and find the social network that is related to the activity in the pilot study. Through an iterative process, researchers can identify how both central and peripheral events and people are connected with each other and associated with the main activity. To avoid missing important details, we suggest applying long-term ethnographic approaches, such as interviews, shadowing, behavior tracing, and self-documents. Recognize participants: in this stage, all people’s goals, roles, interests, and participation in the target activity should be clarified. This information can help researchers to further distinguish different user types and identify potential users. Selection and focus: a workshop or a focus group in this stage can help development teams decide which types of users should become the target audience and to set clear priorities for design development. Once the user types are selected, surveys and in-depth interviews can be used to capture in-depth information about these selected users.
To make the research results serve the same purpose as user models and personas, both dynamic and static information need to be highlighted while portraying the users. First, the portrait (it can be presented as a model or a document) should contain basic demographic information, such as gender, interests, and preferences. However, different from normal user profiling, we highly recommend that researchers also include dynamic user attributes that help predict how users’ behaviors will change with shifts in their demographic data. For example, predictable increases in income are associated with transitions from a teenage life to maturity and result in changes in IT consumption and use. The second type, static information, is about higher-level concerns, including users’ attitudes, motives, beliefs, and their sociocultural background. As mentioned above, this information is permanent and can enrich design solutions and can help evaluate designs in different phases of design cycles and in different projects.
Through contextual design flow modeling, multilevel social activity modeling, and traditional ethnographic techniques, we have shown that there are different types of invisible users, who are involved in social activities, but are neither direct product users nor customers. In the case of a tea ceremony, the young generation, who has inherited tea-drinking habits from their parents, is not considered as a target audience of the traditional tea industry and can only access very limited information. In the case of football watching, potential supporters represent a large group of people who participate in the activities regularly because their children, parents, friends, or colleagues are football fans, and they are willing to share these interests. However, most entertainment and media companies have only considered providing services for football fans. On the basis of what we learn from the three studies, it is argued that discovering all user types and presenting static and dynamic user portraits will bring many long-term benefits for systems and services. Identify potential users: in the three cases, we have shown that extending the research scope can help gain the necessary understanding of complex social interactions and social contexts and discover potential and indirect end users. We consider this to be significant understanding for serious games and technology enhanced education systems. Support diverse and large user groups: development teams can predict new needs and expectations of future target audiences by identifying predictable changes in users’ IT-product usage and economic status. Furthermore, identifying user types and motives can help designers propose more reasonable design solutions, satisfy users’ different needs, and develop more adaptive systems. Support multiple projects: compared to the outputs of usability and activity research, sociocultural information such as attitudes, values, and norms is more permanent in time and is valid across different media and physical spaces. Therefore, this type of knowledge can continuously be reused and reapplied in different projects. For instance, both the first and second studies have shown people’s daily life and online activities are driven by norms and sociocultural values, which can always be considered in social media development.
As discussed in Section
Traditional user modeling mainly focuses on human-computer interaction. Although the traditional models reveal detailed information of the ways that people interact with systems, they are incapable of capturing many factors and contexts critical to design for social interaction, adaptive systems, and serious games. Personas, on the other hand, represent rich information about users’ lifestyles and attitudes and bring various benefits to design practices. However, there is little consensus regarding suitable methodology for how to develop reliable and representative personas. Both user modeling and personas narrow down their audience groups by overlooking activity contexts. User modeling only focuses on direct users of systems, and personas mainly represent target consumers, rather than describing the many participants involved in the activities.
In this research, we compare the strengths and weaknesses of different design and research methods. We, first, present additional types of user information that were gained through application of multiple research methods. Second, we discuss the benefits of having this information in a design process, including being able to identify invisible users, increase awareness of different user types, and develop more informative and representative user portraits. We also suggest suitable research methods in each project phase and list the important information needed to develop user portraits. In addition, we provide many examples through our cases to illustrate how the knowledge of user types can be applied in a design process, in particular in the development of large-scale services, adaptive systems, and serious games. We are aware that the cost of applying long-term ethnographic approaches and multiple modeling strategies to gain these insights is high. However, following scientific approaches can contribute high-quality and reliable user portraits with significant dynamic and static user information, which can later be reused and reapplied in different future projects.