Towards Acceptance Engineering in ICT for Older Adults

2.1 Technology Acceptance Concepts

In the past 30 years, research on technology acceptance has evolved from a psychological view on decision making and user behavior into a multi-disciplinary notion of the relations between system, user, context, and usage. Managers, designers, engineers and advertisers alike aim at tailoring useful, easy-to-use and fun technology for various user groups. To achieve this goal, stakeholders consult different concepts of acceptance, originating predominantly from the development of business software and human behavior prediction.

Venkatesh and colleagues provide a literature review on user acceptance of information technology in the work context, which offers a good overview of existing approaches. In an attempt to unify the view on user acceptance, the authors compared theories and models in accordance to factors that shape acceptance. Consequently, the authors present the Unified Theory of Acceptance and Use of Technology (UTAUT) with the four core determinants of intention and usage: performance expectancy, effort expectancy, social influence and facilitating conditions [30]. Among others, UTAUT is based on the following concepts. The Technology Acceptance Model (TAM) by Davis is still an often used measurement for predicting and explaining system use based on the key concepts perceived usefulness and perceived ease of use [6]. In multiple user studies it has been shown, that external stimuli of system design have an influence on the cognitive response, which includes perceived usefulness and perceived ease of use. These aspects have an impact on the attitude towards using the system, and subsequently, actual user engagement with the system. Davis also points out that the perceived usefulness has a much higher influence on user acceptance than the ease of use [7]. Further theories differentiate the key concepts of user behavior into the attribution to either the user or the (social) context. Examples include the Theory of Planned Behavior (TPB), which is based on the Theory of Reasoned Action, and the Motivational Model (MM). Within the TPB, Ajzen introduces the attitude toward behavior as the emotional reaction of a user, the subjective norm as a socially anticipated action of a user, and the perceived behavioral control as the ease- or difficulty-of-use when performing a task [1]. The MM divides usage motivation, as a core outcome of acceptance, into extrinsic and intrinsic motivation. While intrinsic motivation ascribes the motivation to perform a task to the activity itself, extrinsic motivation promises an outcome or value for performing an activity [28].

While the parameters of acceptance vary when looking into the different theories and models, all concepts show the relevance of an additional value when using technology, e. g. the perceived usefulness, emotional reinforcement, social encouragement, or fun. This aspect is especially relevant, when creating technology for older adults.

2.2 Technology Acceptance by Older Adults

With the purpose of designing successful ICT systems for older adults, user acceptance and motivation need to be taken into account. Since age is considered to play a moderating role in predicting technology acceptance [30], research should discuss and refine the proposed generic models. While the presented user acceptance models explain attitude toward using technology, self-efficacy, and anxiety not to be direct determinants of use and intention, these aspects might have a higher indirect influence on acceptance when focusing on older adults. Multiple studies have shown what factors of the presented technology acceptance theories and models are relevant in ICT for older adults and which additional factors should be considered. Van Biljon and Renaud investigated whether acceptance models are applicable to older mobile phone users. The results of a qualitative study indicate that there are intergenerational differences in the process of adopting technology, and thus, the priority of factors influencing technology acceptance. One major difference is based on the buying decision. Most participants in the study did not buy their mobile phones, but were given the phones by relatives or peers. The authors argue that external mediating factors, such as affection and safety concerns, contribute to raising the actual system usage while not correlating to the perceived usefulness, perceived ease of use and intention to use the system [29]. This contribution provides interesting perspectives in engineering, including the identification of value, not only for the user himself / herself, but for family and peers, as well as the analysis of potential external factors that facilitate or limit technology acceptance. In a study on predicting older adults’ technology acceptance behavior, Wang and colleagues pointed out four contributing factors: needs satisfaction, perceived usability, support availability, and public acceptance. Furthermore, the study results show that support availability was stated to be the second most important factor after needs satisfaction by the older users, whereas measured correlations put perceived usability into this position [31]. While all of the presented factors can be translated into generic user acceptance factors (needs satisfaction vs. perceived usefulness, perceived usability vs. perceived ease of use, public acceptance vs. subjective norm, support availability vs. social influence or facilitating conditions), the conducted survey provides an insight into practical implications when creating ICT for older adults, namely the importance of designing for support.

As stated above, moderators, like demographics or emotional characteristics, may play a different role when designing technology for older adults in comparison to technology in general. The presented studies do not prove a need to change established concepts, but draw attention to practical implications which are not apparent when looking into theories of technology acceptance. Relevant moderating variables have shown to vary according to stakeholders and application context. In literature on ICT for older adults, interaction aspects are often investigated in relation to acceptance. When interacting with technology is perceived as an effort that needs to be made before gaining access to any benefit technology offers, drop-out rates are likely to be high [25]. As indicated in multiple user studies on older adults’ experiences with mobile phones, user satisfaction is affected by perceived usefulness, ease of use, and pleasure of use (Lee 2007). Thus, monotonous interaction and contents or unattractive design turn out to be acceptance barriers. From an expert’s point of view, issues for Smart Home technology acceptance can be grouped into technology design (function, reliability, usability), ethical consideration (privacy, trust, loss of dignity), user perceptions (safety, independence, stigma), and role of markets and policies (access to technology, affordability, different policy) [5]. Perceptions on technology design represent parameters that can directly be influenced by technology design, whereas ethical considerations and user perceptions draw the focus on socio-technological integration. Furthermore, when aiming to create market-ready technologies, financial prerequisites of the user group and policy are additional factors to be taken into account.

Because acceptance is crucial for the success of technology, many approaches tend to encounter challenges for creating novel ICT for older adults. In contrast, there is a high potential in gaining knowledge on the relation of user acceptance and introducing and developing technology for heterogeneous users. Thus, the following chapter summarizes insights into system parameters that have shown to increase technology acceptance in older adults.

3.1 Development of a Cognitive Training Game for Older Adults

In a research and development project, we followed a playful approach with the aim of creating a psycho-motoric training application that helps older adults to remain cognitive abilities [14]. Our motivation in starting this attempt was to react to an increasing need of frequent cognitive training for older adults with mild cognitive impairments and dementia in its initial phase, while less therapists are available. As a tool mainly for therapists but with the user group of older adults, the system was supposed present cognitive training exercises in accordance to user skills. Therefore, dementia tests were digitalized and cognitive training exercises created. We followed the user-centered design process, while iteratively implementing and testing technology. Throughout two years, a game-based android application on a tablet device was developed and evaluated, which introduced cognitive tests and puzzles as well as tasks in a sequential way (see Figure 1). At the beginning of the funded project, we conducted an analysis, deriving requirements from of a literature review, interviews with three therapists and two doctors in a geriatric day clinic, as well as a preliminary study that tested existing tablet applications with the user group. We generated a set of specifications according to the identified needs and potential benefits of technology for older adults. The translation of requirements into specification was made merely on our prior knowledge of software development. At that point, we regarded therapists, peers and relatives to be our main potential customers, while older adults were our target user group. Within the process of creating a system that aimed at providing a benefit for users with cognitive decline and aiming at introducing a product, the required additional value was mainly drawn from doctors’ and therapists’ statements. While therapists were our primary information source, the actual user group of older adults tested existing applications. This approach allowed us to identify issues for people with several impairments, but disregarded the perceived usefulness of potential solutions by the user, leading to potential acceptance issues already at an early phase of the system development. In the design phase, our team came up with numerous concepts on how to support cognitive abilities of older adults. The decision to develop a playful cognitive training game as well as specific decisions within the design phase were mainly based on literature of geriatric therapy, conventional training methods, existing systems for cognitive training, and finally, our own preferences. The only source of preferences with a focus on the users’ perceptions could be drawn from our preliminary user tests, which identified particular content, interaction, and design preferences. Examples include a preference for local information, nature, and sports, and difficulties to perform drag-and-drop tasks [14]. Nevertheless, by failing to discuss likings of our targeted user groups in a participatory approach, more acceptance risks emerged before implementing the first prototype.

Figure 1

Woman using the cognitive training game [14].

After implementing the ideas, we started the first user tests. They already showed a wide range in the users’ attitudes towards the system. On the one hand, some older adults could easily manage to perform all tasks, resulting in few usability issues and a high joy-of-use. On the other hand, the majority of participants was not able to use the system, because interaction and / or content cognitively overwhelmed them. To comprehend the reasons for these issues, we discussed our results with the therapists, resulting in implications for content adaptation. However, only those participants with good cognitive skills, who managed to perform all tasks, were able to communicate their experience with the application in more detail, drawing attention on the validity of interpretation of user study results. Especially when focusing on older adults with various abilities and impairments, our experience suggests that it is questionable whether research results actually reflect the potential technology acceptance. Moreover, discussions that were not part of the conducted questionnaire, but evolved as a side-communication in the testing sessions, proved to be interesting information platforms in respect to perceived usefulness and ease of use. For example, one of the users stated that he was surprised how easy it was for him to use the application on the tablet device, since his children always kept him away from novel technology, claiming it was too complicated. This informally acquired information points out that the social context does play an important role for technology acceptance, confirming prior research on user acceptance in older adults.

At the end of the project, the rollout of the developed system was hindered by several factors, including a lack of resources, and policy for systems that are officially used in therapy, but also because many target users did not perceive the application as being useful, preferring analog cognitive training tasks. Overall, these experiences show the importance to take into account acceptance criteria throughout the whole engineering process. Moreover, we could identify several acceptance aspects when designing ICT for older adults, which we can apply in future projects.

3.2 Development of Movement-based Games for Older Adults

Another set of research projects focused on the development of movement-based games for older adults [9, 10] to support physical activity in care home environments. The goal of the projects was to explore the design of movement-based interaction suitable for older adults who experience a wide range of age-related changes, and to better understand how technology-based support systems can be integrated in care home environments. To this end, we worked with a physical therapist to identify suitable interaction paradigms, and involved nursing staff in discussions around deployment challenges, for example, support that would be available during play, and spatial requirements of the care environment.

In the first project, we followed a user-centered design process that built on previous work on games for older adults to identify suitable game themes. Additionally, we prototyped interaction paradigms, and these were reviewed by a physical therapist for early-stage user testing with the intended audience, older adults living in long-term care. This iterative approach to designing the movement-based interaction allowed us to gain early insights into the suitability of our designs for the target audience, and solicit feedback from nursing staff. Following the prototyping stage, we developed a movement-based gardening game [9] that was in line with findings from the prototyping design, and previous recommendations for the general design of games for older adults in long-term care. Throughout the testing stage of the final game, it quickly became apparent that the design was suitable for the majority of users, but that older adults who experienced severe age-related changes – particularly cognitive impairments – struggled with the combined challenge of learning how to control the game, while also having to understand game mechanics and follow the goal of the game. As this affected their overall gaming experience, this situation created difficulties for the general acceptance and use of the technology in an empowering, enjoyable context.

In a follow-up project [10], we built on these findings regarding the needs of older adults with cognitive impairment, and developed a suite of movement-based casual games that relate to tasks of daily living (e. g., cooking) and were controlled through simple hand movements. Using these games, we carried out a long-term user study at two research sites (a long-term care facility catering to independent older adults, and a care home offering high levels of support), and found that all participants were able to gain access to the games. However, the longitudinal nature of the study revealed two important aspects that need to be considered with respect to acceptability of technology. First, we learned that many older adults were apprehensive to engage with new technologies in a social setting, suggesting that acceptance does not only depend on the technology at hand, but also on contextual factors. In our case, this led to a change in research protocol at the care home that had not initially been planned, but was necessary to accommodate the needs of the participants, and allow them to engage with our games in a safe, encouraging environment. Second, the study revealed huge differences in understanding and progression in the use of technology. While the provided gaming system had long-lasting appeal for many users, others improved quickly and were looking for more challenging tasks, outlining another challenge in the creation of technology for older adults that has long-term appeal.

Summarizing the insights that were gained throughout both projects, they provided a deeper understanding of acceptance engineering of technologies for older adults specifically regarding the deployment context, along with individual user needs that can only be understood through the long-term study of their experience with new technologies.

3.3 Development of Support Means of ICT for Older Adults

A further project focused on ICT to support speech language therapy for older adults [14]. We learned how cooperation of different disciplines changes the view on ICT requirements. Research on how ICT is used in aphasia therapy motivated us to address the needs of older adults. We envisioned an easy-to-use system, which helps older adults to practice their communication skills. Because it is oftentimes embarrassing for older adults with post-stroke speech impairments to communicate in public, we wanted to provide a support in practicing sentences in a protected environment. Therefore, requirements were derived from speech language therapists in interviews with other therapists as well as observation of training sessions, leading to a high application focus. The analysis investigated the usage of therapy means, including hints in speech language therapy, which therapists apply when a patient encounters difficulty to manage a task. Auditory (e. g. repeating a task), visual (e. g. highlighting possible solutions), and tactile assistance measures (e. g., pointing on a picture) were then analyzed with respect to frequency of use. Results indicated that some hints were used more frequent and more successful than others.

The hints were introduced in a specification workshop with three experts in the field of speech language therapy, four human-computer-interaction researchers, and two software developers. The results of the cooperative translation of practical requirements to system specifications had a major impact on the following design decisions. In contrast to our initial goal to create a separate speech language therapy system for older adults, we focused on providing a meta support for existing applications. Several auditory and visual hints were identified that could be integrated not only in digital aphasia therapy, but therapy and rehabilitation applications for older adults in general (e. g. limitation of answers, show an image, highlight or repeat parts of the task). Thus, in the following engineering phase, we were able to focus on how to implement these hints into systems on the market. In a Wizard of Oz experiment, our first sketches were perceived to be useful by both users and therapists, and showed an improved task fulfilment of an aphasia therapy measure in comparison to the same measure without support. Due to the adjustment of an already well-accepted technology, a high acceptance by the user group is anticipated.

3.4 Development of Assistive Transportation ICT for Older Adults

In a different project we looked at the specific transportation needs of older adults [17]. In this project we also engaged in an UCD process, which included the long-term involvement of 19 older participants. The home region of the participants has about 100.000 inhabitants and includes both urban and very rural areas. Public transportation options available are bus and train (mainly for inter-city travelling). The bus service is very limited, especially in rural areas. Additionally, the landscape is very hilly and diverse. The participants were generally in good health, but due to age related impairments they were in or envisioned the transition of from using their own car to using public transport or ridesharing.

We followed an iterative approach, which started with initial interviews to understand the context of older adults’ transportation habits and needs and which was followed by several prototyping and evaluation cycles. We developed multiple paper based and interactive prototypes and iteratively built a mobile application that was tested for 1.5 years. Our findings implied that the acceptance of new, potentially assisting ICT to support the transportation needs of older people is not solely created by solving logistical problem of getting from A to B and taking into account age related impairments. Their choice to make use of ICT to find suitable transportation was mainly driven by two factors. These two factors are ‘independence’ and ‘decisional autonomy’. The way these dimensions are addressed in communicating with the respective users, clearly influences the acceptance of ICT and the implied transportation choices [16].

The findings of the study imply several things with regard to acceptance engineering. Firstly, we found that solving the basic problem of getting from A to B had to be framed in socio-technical context (taking into account personal habits and attitudes as well as social structures). Secondly, our research was embedded in several informal sessions, which resulted serendipitously throughout the studies’ process (e. g. talking about new feature requests during coffee breaks of studies: see figure 2, or at the informal “Christmas party” with the study participants). Thirdly, the introduction of well-designed ICT for older adults has to be accompanied by the introduction of social means of support and feedback. Older adults’ appropriation of technology is characterized by learning and applying instead of trial and error.

Figure 2

A group of older adults participating in a user study.

3.5 Cross-project Insights on Supporting Older Adults Through ICT

The above introduced examples changed the way of how seriously we consider acceptance criteria when starting new ICT projects that aim to support older adults. In addition to these experiences, we can draw findings from our prior research in the field of Ambient Assisted Living, as presented in the following example. In a living lab approach, we invested intensive work on the socio-technological integration. In periodical meetings with the user group of older adults, we discussed usefulness, sense making, privacy, and usability in novel ICT. At the same time, we offered workshops to introduce technology, reduce fears, and provide a protected space, allowing for failure and learning in the interaction. Furthermore, during the entire development process, early scenarios, personas, and prototypes were used to discuss benefits and challenges. Thereby, we experienced that artifacts that are developed within a project, are useful instruments to mediate between developers and users [19].

In summary, our experience in following a user-centered design process is ambivalent. On one side, some attempts to create accepted products were not successful. On the other side, we have gained valuable insights on how to improve our future research, bearing in mind acceptance criteria at all stages of the development process. In the following, we present our approach of acceptance engineering, incorporating lessons learned from our research.

4.1 Lessons Learned from Projects on ICT for Older Adults

Based on our experiences, there are three major sources of acceptance risks: (1) the user himself / herself, in respect to heterogeneous abilities, preferences, and prior experience, (2) the social context, including social structures that may both foster and hinder technology acceptance, and (3) the design team, referring to decisions about which methods to use, the execution of those methods, and personal characteristics. We present our lessons learned, which can be related to one or more of these sources.

Lesson 1: Understand acceptance.

The foremost lesson we had to learn in all projects is that in order to create technologies that reach high acceptance rates, we need to understand the nature of acceptance in relation to the user group, and the application area. With this contribution, we took a first step in analyzing how our experiences can be related to prior research on technology acceptance. However, the objective to fully understand acceptance criteria of older adults is an ongoing process. In any case, reflecting upon one’s own experiences and literature in the field of acceptance, provides impulses towards creating engaging ICT.

Lesson 2: Evaluate the importance of user acceptance for your goals.

When comparing our research experience with existing literature on ICT for older adults, we noticed a difference in the relevance of acceptance, especially in the goals of research and practice. For some research approaches, initial technology acceptance might be essential, e. g. to avoid dropout rates (compare our fourth example). Depending on the business model, products on the market might need to either achieve long-term engagement or persuade customers in comparison to other products (compare the first two examples). Therefore, it is important to evaluate the relevance of different user acceptance criteria for the own objectives and plan methodology accordingly: for some ICT projects, usability may be the core important measurement, but other projects might require long-term user studies in order to evaluate user acceptance.

Lesson 3: Ponder whether it makes sense to continue to pursue your goals.

With the aim to create suitable and well-adopted technology for older adults, it is sometimes necessary to put all thoughts on invested efforts aside and discuss whether goals need to be changed. In many cases, the adjustment of measures and methods – even if it costs time and effort that have already been invested – may provide a benefit in the acceptance outcome. However, any decision in this context should be carefully considered. Our third example showed, that it is not always useful to create yet another application to support older adults, but other measures that improve existing solutions might increase acceptance.

Lesson 4: Always ask users about their wishes, experience, and perceptions of potential technology.

Our first example has shown how technology acceptance is at risk in any phase of the engineering process, when perceptions of the actual users are disregarded. The interpretation of user behavior when interacting with technology is not sufficient to draw requirements from, nor is it a suitable replacement for intensive communication with the user group. Our second example demonstrates that the evaluation context of technology also has potential to reveal vulnerability; in this context, it is important to maintain flexibility and be open to adapting the research process.

Lesson 5: Use validated methodology, but always consider off-the-record information.

Although mixed-method approaches are increasingly used, user studies differ in the applied research design and methodology. But even if off-the-record information is not always helpful, e. g. for quantitative research, this information opens an insight into the perceptions and feelings of a user. Especially for older adults not showing high affinity towards technology, simply talking to people can help discover barriers and motives. Our first and fourth example show how informally gathered information adds to understanding acceptance of older adults.

Lesson 6: Connect the objectively added value with the perceived usefulness of the user group.

From our first and second example we learned that our envisioned benefits did not always meet the perceptions of the user, and the benefits could not always counterbalance the invested efforts. In the analysis of requirements and definition of specifications, it is crucial to acknowledge a potential discrepancy between actual and perceived value, in order to react to it. Communication is a sensible matter and the complexity of user expectations, perceived usefulness, added value, and social context should be analyzed before testing systems with the user group.

Lesson 7: Balance different views on perceived usefulness.

It is not only important to consider objective and subjective usefulness, but to balance a variety of perceptions on this aspect. As a particularly heterogeneous user group, older adults differ in their experiences, abilities, and preferences. Therefore, researchers and developers need to figure out, which viewpoints to focus on or how to include the plurality of perceptions. Furthermore, when considering that older adults are not always the customers of ICT, but their peers and relatives might buy technology for them, novel approaches need to balance customer and user needs.

Lesson 8: Take mediating factors into account.

Even though research on theoretical models points out significant impact of individual determinants on the intention to use a system and the actual usage, it is important to understand that these models tend to abstract the concept of acceptance. These models are intended to be generic, and thus, transferable. However, when developing ICT for older adults, aspects of the application domain and user context should be drawn into focus. This becomes clear when evaluating information on the social influence on acceptance, e. g. family members, who keep older adults away from, or encourage them to use novel technology.

Lesson 9: Use emerging artefacts as common ground for discussing technology.

Throughout the process of developing technology for older adults, many useful artifacts are created, e. g. mind maps, sketches, mockups, user stories, personas, or prototypes. Each of these artifacts can be used to discuss technology with different stakeholders, while establishing common ground. Thus, in the process of generating these artifacts, it is essential to make sure they are understandable for wide audiences.

Lesson 10: Be sensitive to your own actions.

Starting with the requirements analysis and continuing throughout the whole engineering process, the character of the researcher has an influence on the outcome of a user study (see Rosenthal effect [26]). Even if clearly structured, there will always be emotional aspects when directly interacting with users. To a certain extent, these variables can be controlled in order for researchers to become accepted technology providers [20], but one researcher is likely to receive different reactions from the same user in the same situation. Whereas it is not necessarily important to control one’s own actions, the interpretation of user study results should incorporate how one’s own actions effect a user.

Lesson 11: Avoid misinterpretation of results.

When aiming to improve the way we create technology that is both engaging and empowering for older adults, we noticed that not only the different phases of system engineering, but also the transitions in between the phases can have an impact on user acceptance. Transitioning from requirements analysis to specification, from specification to design and from design to implementation is not trivial. In order to not misinterpret the user’s needs, it is advisable to include methods that allow early feedback, such as paper prototypes, Wizard-of-Oz evaluation or focus groups. These measures minimize acceptance risks and strengthen socio-technological integration at the same time.

Lesson 12: Communicate clearly and provide transparent information.

In order to discuss artifacts, goals, and implications, we have experienced a clear communication of results to be a core element for working towards a common vision. However, this requires a clear concept of your own goals and methods, as well as distinct definition of related aspects. Also, when discussing novel technology with older adults, our experience has taught us to provide transparent but easy to understand information in order to establish mutual respect.

These lessons provide an insight on the implications of acceptance criteria on the process of system design. In order to improve our future work, in the following, we show how these lessons can be practically incorporated in the engineering process.

4.2 Including Experiences in Acceptance Engineering

Since implications of technology engineering for user acceptance takes place during every stage during the engineering process, we propose how the above stated lessons can be incorporated into a systematic acceptance engineering process, referring to the stages of the human-centered design process. Figure 3 provides an overview of how to implement the lessons learned in accordance to acceptance engineering. Before starting the requirements analysis, we propose to conduct measures to ensure environment acceptability. In the turn of the process, all phases should be supported by respective acceptability measures (e. g. requirements should be validated by means of requirement acceptability). In the following, we explain these stages and provide examples on what measures to implement.

Figure 3

Acceptance engineering in the human-centered design process.

Before starting to design ICT for older adults, understanding acceptance (Lesson 1) should be an objective in all approaches. This includes a literature review on acceptance criteria for older adults in general and an insight into the application domain. For example, in order to create a system that supports mobility, the environment of the focused user group should be visited. From our experience, the best way to analyze the application domain is to opt for an ethnographic approach, and accompany older adults in their daily routine and observe potential challenges and chances as well as existing practices. Thereby, suitability to the environment can improve acceptability.

This step already contributes to the engineering phase of requirements analysis and can contribute to defining the relevance of acceptance in the own approach (Lesson 2). E. g. in order to create ICT to change movement behavior in older adults with a focus on the neighborhood, long term studies must be conducted that evaluate the behavior before introducing the system, and needs to be continued in several iterations afterwards. We suggest to create a diagram showing the dependency of factors, and to use this diagram as a basic artefact for discussion. Based on this, first materials can be created that structure the requirements analysis (e. g. interview questions). Often, when a system engineer analyzes requirements, he already has a certain solution in mind. In some examples, these have an influence on the analysis itself, or even on early stages of preparing the analysis. Consequently, the personality of the requirements engineer and the relationship with the user play a role in the outcome of this phase. Therefore, it is crucial to reflect on one’s own actions and thoughts (Lesson 10). From our point of view, the simplest way to achieve this is to discuss prepared methods and materials with at least one colleague who is not involved in the project, and to document the outcomes of this discussion. Materials that are prepared in that way can also represent valuable artefacts for discussion (Lesson 9), help to define goals, and provide the basis for clear communication (Lesson 12).

The atmosphere for requirements analysis should be familiar for older adults (e. g. a known community center) to avoid unnecessary stress. Also, we suggest to prepare materials in accordance to the identified system goals, but document the results (e. g. answers of first user interviews) without relating it to the goals. For example, a scenario-based discussion of the application domain is a good measure to agree on terms that are easy-to-understand for all stakeholders. Within this discussion, users’ perceptions should already be integrated (Lesson 4). Different stakeholders should be drawn into this acquisition of information (Lesson 6, Lesson 7). In a focus group in addition to individual interviews, multiple viewpoints can be confronted, providing room for avoiding misunderstanding. From our experience, the different viewpoints should be explicitly stated, and potential overlaps should be identified. Within this discussion, it is important to stay attentive and note possible off-the-record information (Lesson 5), e. g. by recording sessions on camera or assigning one person specifically to noting comments on attitudes and feelings. These measures help to achieve requirements acceptability.

In the phase of translating requirements into specifications, the different artefacts should be incorporated. One possible way to address the needs and preferences of different stakeholders is perspective-taking in discussions. Each system designer prepares the perspective of one or two stakeholders and incorporates his / her arguments in the collective phrasing of specifications. While this process takes a lot of effort, it helps to keep in mind different viewpoints and helps to avoid misinterpretation (Lesson 11). This may also include different perceptions of usefulness and mediating factors (Lesson 7, Lesson 8). The resulting specifications should in turn provide the basis for the discussion with end users. Within this discussion it should always be the goal to communicate clearly (Lesson 11, Lesson 12). A mutual understanding of a concept can best be achieved by asking stakeholders to take the part of the system designer, and explain specifications. Paper prototypes of known systems (e. g. television or even ideas of a service) can be created together to provide common ground. Specifications acceptability provides the basis of an acceptable system design.

In the translation from specifications to design, domain knowledge on how to design for older adults, and how they react to technology is needed, in addition to the basic domain knowledge for the application that is being developed. Therefore, it is necessary to talk to experts in both fields and collaborate closely. For example, when creating ICT for supporting transport for older adults with cognitive impairments, it might be helpful to talk to older adults, bus drivers, as well as neurologists. Until this phase, there is always a high potential that discussions might result in idea change (Lesson 3). At this point, we suggest to discuss requirements and specifications with other community members, and discuss possible solutions before presenting one’s own design ideas (in case disclosure is legitimate).

As soon as all relevant stakeholders agree on one solution, or the benefit of one solution over all other becomes evident, implementation can be started. In this phase, it is important to proceed with rapid iterations (e. g. paper prototypes, Wizard-of-Oz prototypes, click-dummies, or standalone components) in order to conduct multiple user tests and minimize acceptance barriers (Lesson 11). Implementation acceptability or system acceptability also includes user experience and system efficiency and can be evaluated through user tests.

In the phase of user tests, researchers aim to receive an in-depth understanding on the potential usage of the envisioned system. It is important to reflect on results, discuss them with stakeholders, and think about ways to improve the system when focusing on acceptance. When preparing user tests, the socio-technologic introduction should be included. For example, simple tutorials using images and simple wording help users to get accustomed to technology. Also, we experienced the environment to play a major role in the willingness of older adults to test ICT. Our users were always happy to be invited for coffee and cake in a comfortable room rather than sitting in a sterile laboratory. By modifying the test environment, test acceptability can be improved.

Finally, as soon as a system is rolled out, the phase of improvement is not only important for the long term use of a system, but also to maintain positive attitudes towards novel technology: when older adults see that their suggestions are incorporated in the improvement phase, acceptance seems to be higher. Thus, improvement acceptability should be taken into account.

4.3 Implications for Acceptance of ICT for Older Adults

In contrast to the presented theoretical models [6, 30], our experiences support the statements of acceptance engineering approaches [2, 8]. Thus, acceptance does not only involve the users’ individual perceptions and social factors when using the system, but also other stakeholders’ (older adults, family, peers) feelings and attitudes towards what is being created, how it is created, and who created it. Consequently, the process of engineering itself has an influence on technology acceptance. Acceptance criteria should therefore be taken into account at every stage of the development process.

We provided a list of lessons learned and provided practical examples on how to incorporate these into an acceptance engineering lifecycle. By drawing from our own experiences in developing ICT for older adults, we argue that we improve the acceptance of technology for older adults by adapting lessons learned and discussing our own approaches in system engineering. In our future research, we aim to modify acceptance by taking into account our lessons learned at every stage of the design process. Throughout the engineering lifecycles of upcoming projects, we then hope to evaluate different methods to connect them to our list. Even though the implementation of our lessons learned in acceptance engineering is not validated yet, our contribution shows that understanding acceptance criteria requires theoretical models, but improving the actual acceptance of technologies is based on the translation of models and theories into applicable actions. In this paper, we started to summarize lessons learned of user acceptance in relation to the engineering process. Gathering further knowledge on how to transition in between the engineering phases and how to make correct decisions in terms of user acceptance remains a challenge for academia and industry.