there are still actions that are left up to the patient. Typically, patients need to setup a Bluetooth connection, wear the sensor appropriately, and charge the battery. A lack of adherence to telemonitoring actions in eHealth ("an emerging field in the intersection of medical informatics, public health and business, referring to health services and information delivered or enhanced through the Internet and related technologies" [7]) is therefore well documented. For example, Eysenbach [8] states that there "is the observation that in any eHealth trial a substantial proportion of users drop out before completion and stop using the application. [...] for many eHealth trials, in particular those conducted on the Internet and in particular with self-help applications, high dropout rates may be a natural and typical feature." The Economist's Intelligence Unit [6] reported on drop-out rates of 144 mobile health companies: 67% of the users of fitness or mobile health apps that needed to enter data manually, stopped within 6 months. Surprisingly, in the case of automated data capturing, e.g., tracking via sensors, even a higher dropout was reported, namely 74% [22].

To maximize the effect of telemonitoring, it is important patients are motivated (and kept motivated) to adhere to their therapy. Research suggests gamification can have a positive impact on health and well-being when it targets behavioral outcomes [13]. Gamification mechanics might thus be able to motivate patients to adhere to their telemonitoring actions. However, it can be difficult to select appropriate gamification mechanics. To help researchers and designers select suitable gamification mechanics in an eHealth context, we propose a focus group protocol based on an anthropomorphism approach.

In the next section, we will first introduce focus groups, present some background on tailored gamification mechanics, and discuss the tendency of people to anthropomorphize technology. Then we will present the proposed protocol in detail. Before we conclude the paper, we present a brief discussion on user types.

Focus groups, sometimes referred to as group discussions [25], differ from in-depth interviews as most data comes from the interaction between participants. As indicated by [2], "descriptions of a focus group vary but usually include a semistructured session, an informal setting, moderation by a facilitator and possibly a co-facilitator, such as photos." They serve to collect rich and detailed feedback. Based on the group discussions, individual answers become sharped and refined. On the other hand, they might also expose differences between participants that are important to address.

Focus groups create a synergy [28] and have been used before in a gamification context. Sepehr and Head [27] used focus groups to identify key gamification mechanics in ERPsim, which is a gamified system to teach SAP ERP. Fitz-Walter et al. [11] used focus groups to gather input for achievement related design elements.

Gamification harnesses the motivational affordances of gameful experiences to influence psychological outcomes and further behavioral outcomes [12]. Effective gamification is a combination of game design, behavioral economics, motivational psychology, and user experience and user interface design [3]. However, Hamari et al. [12] has shown that gamification is not a panacea. Research has shown that different user groups and personality types necessitate different gamification mechanics [4,19,30]. Where some users may be motivated by competition and leaderboards, other users may be discouraged by competitive elements [26]. Whereas some users may be encouraged by challenges and quests, others may dislike the strong focus on achievement. Hence, there is a need for intelligence with respect to which gamification mechanics are appropriate for specific personality types and user groups. Careful consideration is thus needed to select a set of gamification mechanics. In this work, we build on Marczewski's 52 gamification mechanics [17] to let participants of a focus group select gamification mechanics.

Technology can be hard to grasp, especially for non-technical people. Therefore, some people tend to anthropomorphize technology [1]. This does not only work well in the field of Human-Robot interaction [9]: by attributing human characteristics to complex systems, people can make them more understandable; but users can even consider the interaction with computers social [18]. Not only technology, but also gamification mechanics applied in an eHealth domain can be cumbersome to understand. Similar to Vandenberghe and Slegers [31], we anticipate "anthropomorphism to be an interesting approach for so-called generative techniques that aim to gather insights into people's tacit knowledge and latent needs [32] that may be realized in future products or applications."

Based on this vision, we introduce the concept of 'supercoaches' -as an anthropomorphism for gamification mechanics -who aim to motivate users when they experience issues adhering to their telemonitoring therapy. Using the concept of supercoaches, participants do not need to be familiar with technology, nor with gamification mechanics to discuss motivational strategies that might help overcome the issues the designers try to address. These supercoaches thus serve as an anthropomorphism for the actual gamification mechanics that designers want to implement in their system.

The overall goal of the proposed focus group protocol is to determine: 1) which issues patients experience when they use a telemonitoring platform; 2) which motivational strategies are appropriate and which gamification mechanics are suitable to help patients adhere; and 3) how can we personalize and use the context to optimize the use of these mechanics.

The proposed protocol builds on the Anthropomorphism approach as proposed by Vandenberghe and Slegers [31]. Our protocol thus starts from the idea that users are tempted to humanize technology and software. This method introduces the metaphor of an all-knowing, omnipotent virtual 'supercoach' who can, and wants to, do everything to motivate patients. Thanks to this abstraction, patients might be stimulated to think about the future and ideal use of the telemonitoring platform, without having full insights into sensor or gamification mechanics, or personalization techniques such as recommender algorithms.

In each step of the focus group different cards are used: 1) the first card is used to list the issues participants experience when they are using the platform under evaluation (Figure 1A); 2) in the second step, a supercoach is introduced to help overcome the issue listed on the first card (Figure 1B); 3) feedback from the other group (patients vs. caregivers) is added using a third card (Figure 1C); 4) finally, this card trail is extended with gamification mechanics heavily based on Marczewski's gamification inspiration cards [16]. However, for each mechanic we translated and adapted the text to the local language and searched for an appropriate example (see Figure 2 -for this paper, the original English text is used). In this focus group study, we want to determine how [application name] can motivate you to execute all operations and continue to perform them accurately. Over the next two hours, we will brainstorm on this topic using a specific method. Imagine there is a supercoach at your disposal, an all-knowing, omnipotent supercoach who can always motivate you at the right time and who knows perfectly what to say and what to do. This supercoach wants to make it easier for you to perform your actions.

Time: 15 minutes (5 minutes writing, 10 minutes discussion) Setup and Explanation: Provide sample cards (see Figure 1A) participants can use to list potential issues they experienced. You might already know [application name], so you might be familiar with the actions you need to do. In this first step, we want to determine a list of issues where you have difficulties with or where you want more information on. Possible questions are:

• Which actions do you find confusing?

• Which actions do you forget to apply? • Which actions do you not understand?

• Which actions do you find annoying?

• Which actions have you already deliberately skipped?

We would like you to describe and explain this in a short sentence. Also try to provide additional information, where, when, with who, and why? So please do not just write "blood pressure" but elaborate: "when I come home and want to feed the dog, I find it hard to measure my blood pressure slowly." Examples:

• Yesterday, I forgot to measure my blood pressure. This often happens when I have to work late. Examples:

• Supercoach can explain that '97%' represents an approximation of the amount of oxygen in your blood. He can also tell you that normal values fall between 95% and 100%.

• Supercoach can send you a message to measure your blood pressure when you are behind the TV and are sitting still anyway. • Supercoach advises you not to watch the screen of the device anymore, so you do not focus on your heart rate. • Supercoach can show you a screen with the timings you have measured your blood pressure.

Time: 15 minutes (5 minutes writing, 10 minutes discussion) Setup and Explanation: Ask feedback (see Figure 1C) on the supercoach from the other group (patients vs. caregivers).

We are now going to split the caregivers and patients temporarily into two groups. We ask to give the supercoaches to the other group so that you continue working on the other group's supercoaches. Take a card and think how the supercoach's response might look like for you. When, how and in what situation would you expect an answer? Is there anyone with you? Attach this new card to the supercoach.

Examples:

• A video can be too stimulating and affect the results.

• Beware of the nocebo effect by showing too much information.

Time: 30 minutes (12 minutes explanation of gamification mechanics, 11 minutes writing, 7 minutes discussion) Setup and Explanation: Introduce the gamification mechanics and allow participants to choose a gamification mechanic to attach to the card trail (as shown in Figure 1).

We are now moving to the penultimate phase. We will show you some mechanics that might motivate [application name] users. We are asking you to take some of these cards with proposals and then paste them on the card trail.

Examples:

• Earn a badge when you go to sleep on time.

• Earn experience points when you measure your blood pressure at a certain time. • Show a leaderboard that shows how many users recorded their measurements correctly this week. • Show a visualization that shows how well you are doing.

• Provide a clear goal in which you try to use the activity trackers at least during the weekdays.

Time: 30 minutes (10 minutes dot-voting, 20 minutes discussion) Setup and Explanation: Dot-voting.

We now have many supercoaches full of comments and ideas.

In this final step, we will sort these supercoaches. Please look at all the supercoaches and observe what has been written down. Everyone gets five stickers that you can divide over what you think is the most relevant issue and solution. You also get one veto sticker for something you really do not want. On a veto sticker, you will need to write your initials or name so that we can ask you for more explanation. You will get ten minutes to do this, after which we will discuss this further.

Although this proposed focus group protocol might help to select an initial set of gamification mechanics, there is still a need for added intelligence to adapt to user profiles [29]. Moreover, there are contexts in which users may be less inclined or simply unable to follow up on certain gamification mechanics. It simply might not be the right place or time to follow up on a challenge or start a competition.

There are multiple frameworks available that help to determine the users' type [3,14,24]. A well-known example is the Hexad framework [15] of which Tondello et al. [30] "created a 24-items survey response scale to score users' preferences towards the six different motivations in the Hexad framework.".

Marczewski [15] suggests different game mechanics that may support different user types. Orji et al. [20] "reveal that people's personality traits play a significant role in the perceived persuasiveness of different strategies."

A potential extension to our protocol is to measure the participants personality. Participants might be asked to complete a short Big Five test [23]. In the focus group, this test could be a survey. However, in the eventual implementation this could be automated. For example, by linking social media profiles of the user [10].

In this position paper, we proposed a protocol that can be used to select suitable gamification mechanics in an eHealth context. We make use of the tendency of people to anthropomorphize technology, which makes it easier for participants to brainstorm on issues and, more importantly, on motivational design techniques such as gamification mechanics.

We are scheduling multiple focus groups with pain patients who are asked to use a telemonitoring platform for 10 weeks after their operation. Although we focus on telemonitoring actions in our work, we do see an added value of our work for the related well-being and flourishing domains. For example, it might be hard for people to report on gamification mechanics that might motivate them to become more autonomous or pursue additional personal growth. However, using this proposed 'supercoach' protocol, participants do not need to be aware of these gamification mechanics before they can start the discussion.