When creating the contextual personas, the healthy work model by Karasek and Theorell  was used. Hence the first section below presents this model. This is followed by a presentation of future offices and the emerging trends in this area. Finally, we present a short description of research on the personas method.
2.1 Digital Work Environment and Healthy Work
As work has become more digitalised the digital work environment becomes more central for employees productivity and also wellbeing. The digital work environment is here defined broadly as the work environment that is the result of digital tools and work support systems. The digital work environment consists of physical, psychosocial and cognitive problems and challenges. It encompasses all cognitive work situations where people interact or are dependent on computer systems. In this definition of the digital work environment we also include aspects of integrity, control and surveillance.
When analysing the data from the contextual interviews, the Demand-Control-Support model is used. In the 1970’s Robert Karasek developed a model for analysing work-related stressors associated with cardiovascular illness. His demand and control model was thereafter further developed together with Töres Theorell  and is now one of the most widely used models for explaining psycho-social work conditions and their effects on health. This model suggests that the combination of perceived demands and perceived control at work is a determining factor for stress. This model was used since our previous research, see for example [2, 3] that have shown that the model is easily understood and applied in organisations when the digital work environment has been discusses. The figure below illustrates the Demand-Control-Support Model (Fig. 1).
The figure above illustrates how healthy and sustainable a work is, in relation to the experienced demands, control and social support. High demands are normally not a problem, if combined with high self-control over work situation and tools and strong social support from management and colleagues. A skilled worker can experience this as a challenging situation. She has full control over the work conditions and planning and gets full support when needed. The work is efficient and sustainable. On the other hand, if high demands are not met by strong self-control and social support, the situation will soon become dangerous health wise. If the worker does not have control over work conditions and planning, does not have usable tools and feels totally exposed if things go wrong, the work will be very unhealthy. Such work situations are associated with high stress. In this extreme, different health risks are common and people do not withstand the situation for long. Research shows that subjective control and support factors often decrease when new IT systems are introduced, as did Åborg, [4, 5].
In this research project we have focused on administrative work. The main health problems in such computer-supported work situations are: Users are bound to use the computer for a major part of their working hours entailing constrained, static work postures for long periods. The computer controls the work pace and task order, leaving the users little or no control over their work. Users suffer from stress, caused by excessive workload, time pressure and poorly designed computer support [6, 7].
Traditionally, occupational health experts work in isolation from the software development process. They evaluate and suggest improvements to existing workplaces and tools . It is, however, often too late to do something about poorly designed software tools once they have been installed and are running. Thus, poor and inadequate design leading to health problems cannot be sufficiently modified post-hoc of systems development when the computer system is in use. Instead, occupational health and ergonomics experts must be involved in the actual software development process. Work-related stress has increased in the past years and since long it is well known that it is a growing health problem . Organisation of work organisation and work content are important factors underlying stress problems, and in work, IT support systems, especially computer software, play a major role. The mental workload tends to increase when new IT systems are introduced , and the decision latitude is lower for extensive computer users than for others .
2.2 Future Offices and its Relevance to HCI
The physical work environment in all its variety is of course another factor influencing office workers . Recently, flexible offices and more specifically management philosophies such as the idea of activity-based working (ABW) has attracted the interest of both organisations and researchers . Place is here seen as a mediating factor between people and IT. Indeed, the strategic use of corporate space is seen as the necessary, though not sufficient, factor in empowering the workforce and ameliorating many of the downsides of computer-supported work . A part of the concept is increased use of IT in support of both mobility and monitoring, and not surprisingly, the IT industry itself is one of the major proponents of this new way of work . Yet, while ABW is proposed as a solution to the problems associated with open plan offices, not least cognitive stress, research is inconclusive . In a seminal paper  traces the origins of this seemingly new–anywhere, anyplace–work rhetoric and exposes some its inherent paradoxes, not least how these images of newness contribute to the conservation of old work patterns.
As IT in the workplace thus becomes more and more embedded and pervasive the scope for HCI broadens. The pioneering works of  have already argued for the application of usability concepts in the field of facility management, yet the field seems to have attracted limited attention so far . Instead, the most common tool seems to be variations on the model developed by ), where the amount of face-to-face interaction is contrasted with the amount of job autonomy (resulting in a matrix of four basic office types: the hive, the cell, the den and the club). While this broad categorisation of work can aid planning, other techniques, such as personas, could provide a deeper understanding on how to improve the quality of work and the work environment.
Within development of IT-based systems, the persona method has become frequently applied tool and is used extensively in both industry and in research. The persona method is a user-centred way of representing users in situations where users cannot be available; the idea is that the overall focus and awareness of the users in development projects are heightened when working with personas . However the use of the method can be manifold for example, the personas are described both as a communication tool and as a design aid. However,  argues that by trying to separate the different ways to use the method will help the usability practitioner to more skillfully use the personas, as well as being open for alternative applications.
Within the field of Human-Computer Interaction, the persona method was originally introduced by Cooper , and he argued for hypothetical archetypes of real users in order to avoid designing systems that become too generic and in the end does not fit anyone. According to Cooper , the personas should be based on actual users and the personas should be precise and specific since it is more difficult to ignore a detailed persona than aggregated user data. The ideas is that numerous personas initially are created through an iterative process, and then these are condensed, according to their goals, into fewer but more precise, personas. One more important claim from Cooper is that even though multiple personas can be created, the developers should focus on one primary persona .
The extensive use of personas result in different views on what should be the basis of the personas. The most common argument is for collecting qualitative data through for example interviews and observations of real users [21–23]. However, the data underpinning the personas does not need to be based on ethnographic studies of real users; for example  describe personas within secure systems design based on assumptions. Quantitative data from for example surveys can also be used to statistically render personas, although these can later be refined by interviews and observations . Moreover, the widespread usage of the method in disparate settings and contexts has made the resulting personas in different shapes and forms.  outline in their paper a loose typology of personas, however, the persona kinds that are described in their paper are not exhaustive; for example, other possibilities are assumption personas  or a collaboration persona . The persona method is also criticised, where the most alarming critique is that the personas are being misused and that this leads to designers distancing themselves from real users .  argues that it is better to engage with users directly than to create a façade of user-centeredness . Other types of critique are that the method is difficult to verify as (more) beneficial compared to other method  or that it is inevitable that designers will create stereotypes . Furthermore there is literature that shows when personas have failed to work, such as the case that  present. The reasons for this was mainly because the design team was not familiar enough with the method and the interaction designers were not involved in the creation of the personas.  present a case where the persona method was abandoned in the context of developing software for mass-market mobile devices, although this mainly was because of the power and dominion of stakeholders outside the development organisation. Some recent studies have shown that the probability of personas being used is higher if the designer has participated in the creation of them [27, 33]. (Furthermore,  presents a case where personas were used outside the development project in which they were developed. In their case the educational department adopted the personas as a way to introduce newly employed to different clusters of customers.