Fluidity in the Phased Framework of Technology Acceptance: A Case Study to Understand (Older Adult) Participant Journeys Through Acceptance Phases with Mobile Telepresence Robots

Abstract

Loneliness has a direct impact on mental and physical health. This is especially relevant to older adults. In prior studies, socially isolated older adults wanted technology that would help them feel more physically present even across distances, such as telepresence robots. However, how useful this technology can be directly depends on whether people accept it over the long term. In this paper, we describe a case study in which we introduced telepresence robots into homes of older adults for seven months. We investigate how older adults’ progression through acceptance phases ebbed and flowed. We describe primary factors that affected speed of progression through acceptance phases: solving problems with technology, life situations (business vs. routines), and personality. We introduce example personas based on this case study. We also propose changes to the longitudinal technology-acceptance framework to take this more nuanced view into account. These outcomes will help future researchers and practitioners to better understand and influence longitudinal technology acceptance.

1. Introduction

Technology can help people maintain lasting and sturdy connections with others. Technology has also made connections across distances more immediate and similar to those in real life, by including more mediums (e.g., audio, visual). In doing so, it can support more natural social connections. Older adults (age 65 and above) have seen socially connective technology evolve from pen-and-paper letters to phone calls to video calls, and most recently, mobile telepresence robots.

Mobile telepresence robots are movable robots that make video conferencing versatile [1]. People operating the robots at a distance, or “pilots” [2,3,4], can call into people housing the robots, or “hosts”. Through the robot, they can experience a two-way video and audio connection as they drive the robot in open spaces that are clear of barriers. Because these robots allow pilots to physically move around their environments, they increase perceived presence in social interactions, and people can gain more benefits (e.g., perceived social support, reduced loneliness) from interacting with loved ones through this medium [5,6,7,8]. However, to do so, users (pilots and hosts) need to be able to accept and use them over the long term.

In this paper, we examined host participants’ acceptance of telepresence robots through seven acceptance phases in the phased framework of technology acceptance (PFTA; [9]). We ran a seven-month study of older adults using mobile telepresence robots for social connection with loved ones, and our prior papers describe participants’ use, frustrations, and recommendations for the robots overall [10] and how participants on average progress through the acceptance phases [11]. However, we found that progression was more nuanced than prior studies reported. Therefore, in this paper, we more precisely examine how each of the four older adults moved through the acceptance phases and how researchers can more accurately determine when a given participant has entered a specific acceptance phase. We study what characteristics of participants, and what contexts of use, aid and hinder older adults in accepting these mobile telepresence robots in the long term. Results will help researchers and practitioners help users move towards a more complete acceptance of technology that has great potential to benefit them.

1.1. Loneliness and Older Adults

Loneliness is an increasing circumstance in our world. Researchers define loneliness as low-quality and low satisfaction with social interactions and relationships [12]. Individuals experiencing loneliness [13,14,15], low levels of social support, and fewer relationships [12,16,17,18] tend to experience increased illness and mortality.

Older adults, aged 65+ [19,20,21], are more likely to feel lonely [22] due to the frequent loss of social relationships, health issues, and mobility difficulties, than younger adults [23]. Older adults tend to travel less frequently and for shorter distances than younger adults, often because of the stress of travel, carrying luggage, difficulties of car operation, and walking distances [24]. Thus, older adults are more susceptible to loneliness due to their limited social ties and interactions outside of their homes [25,26].

1.2. Mobile Telepresence Robots

Mobile telepresence robots can help people socially connect with each other when they are distant. People use these robots to socially connect in entertainment, medical, military, research, and education contexts [27]. Telepresence robots allow people from different locations to be virtually present, like in video conferencing, by allowing users to communicate with family members and friends through a more physical and emotional interaction than with video or phone calls [1,25,28]. Middle and older adults who were isolated from social connections requested access to technology like mobile telepresence robots to help them feel more connected to loved ones [29,30]. By helping people socially connect in this deeper interaction, these robots can help people decrease feelings of loneliness and depression, and improve health and social connection [5,6,7,8].

However, research on mobile telepresence robots primarily for social connection is limited. Some studies examined these robots for social connection in eldercare facilities [25,31]. Residents in eldercare facilities who used the robots felt that the robots helped them connect with their families and decreased feelings of loneliness [7]. Two prior studies examined mobile telepresence robots in homes of independent-living older adults primarily for social connection [25]. In a two-day study, older adults received daily calls via the robot from the research team and up to two additional daily calls from a family member or friend trained to use the robot [5]. Having someone check in on them through the robot reduced loneliness and increased the perceived presence of the caller [5]. In a twelve-month case study, when two participants could use a robot in their homes whenever they liked, they showed continued use of and interest in the robot for social connection [6]. More research should examine the long-term (≥six months) use of robots in older adults’ homes because use and acceptance can change over time [9]; see Section 1.4.

Some drawbacks to telepresence robots include technical difficulties and privacy. Some features (e.g., audibility) can diminish due to poor internet [7,28,32]. Older adults also had privacy concerns (e.g., that a pilot or attacker could find sensitive data like bank information; [33]). See [10] for our discussion of technical and privacy concerns in this study. Other difficulties of the robots relate to general factors in technology acceptance; see Section 1.3.

1.3. Factors Affecting Technology Acceptance

It is important to study user acceptance of technology to better understand what aids and hinders users in accepting and using technology. Many models describe user perceptions and characteristics that affect technology acceptance (see [34,35] for reviews): The Theory of Reasoned Action (TRA; [36]), developed for sociological and psychological users, emphasized attitudes, social norms, and intentions. From the TRA emerged the Theory of Planned Behavior (TPB; [37,38]) to account for the availability of resources, opportunities, and skills, and their perceived significance. The Social Cognitive Theory (SCT; [39]) extends focus to how behavior, personal issues (e.g., personality), and the environment affect acceptance. The Technology Acceptance Model (TAM; [40]) summarizes earlier models, suggesting that perceived ease of use (i.e., it is easy to use) and perceived usefulness (i.e., it has a purpose) are the two primary factors in determining one’s intentions to use technology [40,41,42,43]. The Extended Technology Acceptance Model (ETAM) includes user characteristics, like computer literacy, technical support, and online learning anxiety [42,44]. The Unified Theory of Acceptance and Use of Technology (UTAUT) supports impacts of acceptance including three impacts for the intention of use: performance expectancy, effort expectancy, and social influence and facilitated conditions [45]. The information system (IS) and information technology (IT) Acceptance and Use (IS/IT) model is an extension of the original UTAUT model to include attitude as it was impacted by social influence and had a direct impact on usage [46].

Approaching technology acceptance from another angle, age, research on the “digital divide” indicates that many older adults have less skill, expertise, and affinity for new technologies than younger adults, which can decrease the benefits that technology could have for social interaction [47]. Especially during times of stress, older adults were more likely to revert to using older, more familiar, technology, like telephones (landlines and mobile) or writing letters, rather than social media or video calls [48]. Thus, although mobile telepresence robots may particularly help older adults, using the technology can be especially challenging for them [49]. However, many older adults are willing to learn how to use new technology when it helps them be more independent [50]. The Senior Technology Acceptance and Adoption Model (STAM) includes key factors that can influence older adults’ acceptance of technology, including User Context, Perceived Usefulness, Intention to Use, Experimentation and Exploration, Ease of Learning and Use, Confirmed Usefulness, Actual Use [51].

1.4. Acceptance Phases for Long-Term Technology Acceptance

To truly understand how people think and feel about technology, we must examine long-term use. After two months, researchers can observe ordinary use beyond the novelty effect [52]. After six months, most users have gone through all acceptance phases, and become fully acquainted with, and understand the technology. It becomes more than an accessory: it becomes a personal object through emotional attachment [9].

There are many models of long-term technology acceptance. Some models more generally describe pre-acceptance (before using the technology), acceptance (early technology use), and post-acceptance (after initial acceptance (e.g., [53])). Other models include more detailed descriptions of acceptance. For example, the theory of domestication [54] explains the processes with which users give meaning and significance to the technology. The Technology Utilization Theory (TUT) is based on the importance of how technology is used and if it is being fully utilized by users, focusing on Post Usage [53]. The theory of diffusion of innovations [55] explains how users can reinvent or change technology while using it and the organizational process of structuring how technology and social structures change over time. It focuses on the technology’s characteristics and environmental aspects rather than predictions of outcomes compared to other acceptance models [35,56,57]. The Domestic Robot Ecology (DRE) proposes physical and social space, social actors, and intended tasks to be influential factors of acceptance [58].

In this paper, we used the Phased Framework of Technology Acceptance (PFTA; [9]) because it builds on prior frameworks (Theory of Domestication, Diffusion of Innovations, and TAM) and long-term study findings of acceptance of technology in participants’ homes [52,58,59]. The PFTA has six phases of accepting interactive technology: Expectation, Encounter, Adoption, Adaptation, Integration, and Identification. A seventh phase, Non Use, describes when participants abandon use before or after full acceptance [9]. de Graaf et al. [9] describes each phase and an approximate timeline (

Table 1. Approximate phases timeline according to a phased framework [9].

Acceptance Phase	Timeline
Expectation	Two weeks prior to the introduction of the technology.
Encounter	Introduce the technology.
Adoption	Two weeks after the introduction of the technology.
Adaptation	One month after the introduction of the technology.
Integration	Two months after the introduction of the technology.
Identification	Six months after the introduction of the technology.
Non Use	Can occur at any time.

); however, they recognize that not all users reach acceptance phases at the same time, and some users get stuck in Non Use and stop progressing. These acceptance phases are useful in monitoring the progression, digression, and stagnation of a participant’s acceptance of technology. We describe them in more detail below. To our knowledge, no other studies have used de Graaf’s entire PFTA six acceptance phases [9].

1.4.1. Expectation Phase

In the Expectation phase, individuals have no personal knowledge of, or experience interacting with, the technology [59]. Individuals learn about the technology through word of mouth or research and begin forming an attitude towards it before encountering it [52,58]. Prior studies found that potential users begin forming expectations by asking questions such as, “How does it work?” and “Why does it work?” to seek an understanding of the uses, controls, and functionality of the technology [55]. Initial concerns or uncertainty about usage and privacy can arise. People who are more hesitant about this technology may seek reinforcement to persuade them to move forward to the next phase of using the technology for themselves [56]. Individuals often rationalize using the technology and begin to form attitudes and emotions toward it [9,55]. This phase directly connects to people’s understanding of and preparation for using the technology.

1.4.2. Encounter Phase

In the Encounter phase, people use the technology for the first time. This is the first time they can reevaluate their past expectations of the technology based on their own experience with it and consider new expectations [9,60]. Some people do not advance to the next phase, sometimes because their expectations do not match with the actual technology and its uses, thus causing an expectation gap [61]. People can ask questions, try the controls, and figure out basic uses for the technology in their daily routines.

1.4.3. Adoption Phase

In the Adoption phase, people have had the opportunity to interact with the technology for approximately two weeks [9,55,59]. This causes either excitement (positive experience) or frustration (negative experience). When excited about technology, people tend to gain interest in it and embrace positive feelings or even potential emotional attachments. When frustrated, people either work out their cause of frustration and resolve it, leading to adopting the innovation, or fail to resolve the frustration and reject the technology. There is an important distinction between adoption—the decision to start using the technology—and adaptation (the next phase)—when people incorporate the technology into everyday use [9]. In the Adoption phase, people still have apprehensions and concerns as they try to understand the technology fully, but people are slowly becoming more comfortable.

1.4.4. Adaptation Phase

The Adaptation phase begins as soon as people fully adopt the technology and make the decision to use it [9], after about one month of use, when people know more about their technology [58,62]. People tend to get more curious about the technology and comfortable with it by learning about it, enhancing its uses, and adapting it to them—or adapting themselves to its current use. This phase centers around people understanding their excitement and frustrations, and seeing if the technology will fit into their lives.

1.4.5. Integration Phase

The Integration phase tends to occur two months after beginning the use of the technology [59]. Users have become comfortable with the technology and found different ways to incorporate it into their daily lives [55]. People know enough about technology to enhance activities that they enjoy and avoid or change activities that frustrate them. People may use the technology differently than the designers originally intended [54,63]. Users may incorporate the technology into daily routines and even depend on it [59].

1.4.6. Identification Phase

The final phase, Identification, typically occurs six months after beginning the use of the technology [59], when the technology has become part of their daily routines. It either connects them to others through a sense of community or separates them [59]. The technology exceeds functional purposes and relates to their self-identification, or users form an emotional attachment to it [9]. This is also seen in the Adoption phase, but now it is deeper related to personalization of technology settings, and daily uses that increase and strengthen identification during continued use. Individuals may seek reinforcement of their original adoption and possibly even reverse their decision if negative experiences begin occurring [56]. This creates an intersection of use where people will either continue to use even though they had negative experiences or allow the negative experiences to stop their use. If they continue to use it, they are often comfortable with sharing supportive knowledge about this technology to influence others to use it [9].

1.4.7. Non Use Phase

Non Use—not a typical phase, but an oppositional path of acceptance—occurs if people reject the technology [55,64]. The Non Use phase and codes are not new to literature but new in their relation to the PFTA [9]. Non Use can occur at any time after beginning to use the technology [55,65]. This can occur as a Lack of Use, Non Acceptance, or Suspension of Use. Lack of Use means that users are not using the technology to its full capabilities. This unwillingness to take risks can be related to a lack of experience with technology (new or old; [57]). Non Acceptance is when people contemplate accepting the technology, but after some use, decide to forgo it. An important factor related to a Non Acceptance is knowledge barriers that users experience [66]. Suspension of Use is when a user fully deserts any use of the technology after experiencing all of its capabilities [55].

1.5. Current Study

In our study, we examine how older adults progress through the acceptance stages with a mobile telepresence robot. We follow the PFTA [9] due to its relevance, incorporation of several models, and use for studying long-term acceptance of robots. We use data from our prior study in which we placed a mobile telepresence robot into the homes of participants (aged 60+) for seven months [10]. We interviewed participants monthly and encouraged them to use the robot at their own convenience throughout the seven months. We previously discussed a case study of four participants [11] following the PFTA [9] and the additional Non Use phase [55,57,64,65,66], and found that progression through acceptance phases was much more fluid than prior research suggests. In the present paper, we examine this fluidity in greater depth, including what hinders and aids progression through the acceptance phases. We separately describe how each participant progressed through the acceptance phases and conduct a deeper analysis of the importance of the differences between each phase. Our main research questions were:

• What backgrounds and characteristics help or hinder people moving longitudinally through acceptance phases?
• Can we identify personas to help future scholars assist people moving through acceptance phases?
• How can scholars identify which acceptance phase is occurring?

2. Method

In this case study, we used data from a prior study [10]. The original study included seven participants who were hosts of telepresence robots in their homes. Of these seven, we chose four participants for this paper, due to their diverse experiences and background. Kelly (Sasha, Kelly, David, and Jessica are pseudonyms assigned to participants to protect anonymity) and David had high levels of past technology use and experience; Sasha and Jessica had low or no past experience. Some of these participants experienced acceptance very early, some later, and one not at all.

Each participant received a mobile telepresence robot. The robots belonged to these participants for the entire seven-month study, and they could use the robots in any way they wished. We chose seven months to observe how people interact with the robots through all six acceptance phases, after which the way people treat robots becomes more stable [9]. Participants met monthly with the research team and a family member or friend who piloted the robot to complete an activity and interview. Please see our prior paper [10] for the complete method section. We summarize relevant details here.

2.1. Participants

We recruited participants aged 60+ through public advertisements and word of mouth. We originally tried recruiting adults aged 65+, but due to difficulty in recruitment, we expanded to ages 60+ to recruit our final participants. Prior work sometimes defines ages 55+ as older adults [18,29,30,67]. Participants each recruited one interaction partner (IP; aged 18+)—a family member or friend who lived in a different house than the participant. This case study includes four participants and five IPs (see Appendix A; David could not find a single IP who could attend every monthly interview and therefore switched between two IPs). Participants and IPs were monetarily compensated for their time.

2.2. Mobile Telepresence Robot (Double)

We used a commercial robot: the Double 3 Mobile Telepresence Robot (Figure 1;CA, USA). Double 3 is a 25-pound (11 kg), two-wheeled videoconferencing robot [68]. It autonomously avoids obstacles to help inexperienced pilots navigate the remote environment safely. Double 3’s two cameras allow pilots to see the environment and six microphones help pilots hear people from far away and reduce background noise. It uses Wi-Fi and can function for up to 4 h before charging. It can raise and lower the screen to heights between 47 and 60 inches (119 cm and 152 cm, respectively) [68]. Pilots can call in using a single-use guest access code, which is valid for 24 h, or log in with the hosts’ username and password at any time.

2.3. Procedure

During the seven-month study, we conducted interviews about participants’ use and acceptance of the mobile telepresence robots (

Table 2. Tasks of participants (P) over seven months.

Element	Timeline	Description	# of Times per P
House Tour	2 weeks before introducing the robot	Assess if the home can hold the robot (e.g., clear walkways, stable internet; 30 min); interview (30 min).	1
Bringing robots	Introduction of the robot.	Bring the robot to participant homes. Set up the robot; interview (60 min).	1
Main sessions	Months 1–7	Monthly shared activity with participants and IPs (20 min). Post-activity interview (40 min).	7
Final Session	2–4 weeks after the 7th monthly session	Remove robots; interview (60 min)	1

). We first toured participants’ homes to determine if the robot could drive and operate in them. Then we brought the robots into the homes. At monthly study sessions, we observed participants and IPs doing shared activities via the robot. After seven months, we removed the robots from homes. During each visit, we interviewed participants; see Section 2.4.

2.3.1. House Tours

Before bringing the robots, we assessed participants’ homes to ensure that the robots could function effectively (e.g., sufficient internet bandwidth, and clear paths the robot could travel). We interviewed participants about their expectations of the robots.

2.3.2. Bringing Robot

Two weeks later, we brought the robots to participants’ homes. For each household, all full-time residents attended this session, and IPs joined via Zoom video conferencing. We provided standardized introductions on using the robot. This step was necessary because how a robot is first introduced directly affects its usage [52,69]. We asked participants to examine the robot’s box as if they were buying it at a store and then helped them assemble it. We assisted IPs in switching from Zoom to the robot and learning to use it, its capabilities, and applications for use. We interviewed participants and IPs on Zoom to discuss their first impressions and overall experience with the robot and answer any final questions about the robot or the study.

2.3.3. Main Sessions

Main sessions occurred monthly over seven months. During this, participants and IPs did a shared activity (20 min) together via the robot, with a researcher present in the home. Participants chose a shared activity either from examples the researchers provided (e.g., conversations, cooking, a walk, chores) or that they devised (e.g., book club, decorating for the holidays, board games). These activities differed from typical video conferencing in that pilots, via the robot, could freely move around the house during calls [70]. When participants were simply conversing, they sometimes moved between rooms, like to show each other something, and other times used little or none of the mobility aspect of the robot. Next, IPs signed off from the robots, and we interviewed participants (40 min).

2.3.4. Final Session

After seven months, researchers removed the robots from participants’ homes. Researchers conducted a final in-depth interview via Zoom about their experiences throughout the study.

2.4. Measures

Interviews

We audio and video recorded semi-structured interviews across 8 months. The interviewer was in-person, and the note-taker was present via video conferencing (e.g., Zoom). We interviewed participants before and after bringing the robots, each month after the activity, and at the end of the study. We modified questions from a prior study [71] to explore several topics (See Appendix B for the complete list of questions). Some examples are:

• Use—“How often do you use the robot on average per day/week?” “What activities did you use the robot for?”
• Advantages and disadvantages of the robot—“What are some benefits/disadvantages to this robot?” “How can the robot be improved?”

We followed the phase code framework of [9] with the addition of the Non Use phase for each interview to learn how participants experienced acceptance phases with the mobile telepresence robot; see Appendix C. We applied these codes to our interviews and did not alter or add any questions to better fit this framework or the understanding of the acceptance of technology.

We followed a three-phase thematic coding process: (1) two researchers read the corrected transcripts without coding, (2) researchers reviewed the Phased Framework codes [9] and talked together to clarify the codes for themselves [30,72]. Researchers each coded 41% of the House Tour, 100% of the Bringing Robot, 49% of the Monthly, and 42% of the Final Interview data on their own. While reviewing these codes together, researchers made any revisions necessary to better clarify operational definitions of each code; Appendix C details the phases, corresponding code, and examples of codes for clarity. (3) Researchers independently re-coded the transcripts [30,73] in totality. Each participant’s interview response was assigned a single code; the inter-rater reliability calculated using Cohen’s kappa was 100%.

3. Results

3.1. Definitions

We refer to phases using bold lettering and the phase number subscript (e.g., Expectation₁). For codes within phases, we use italics and the phase number subscript (e.g., Recognize Disadvantages₆).

Prior literature did not operationally define experiencing a phase based on qualitative data. We propose a set of definitions about experiencing a phase and characteristics of participants’ experiences to help guide the reader through our findings. For each interview, we graphed what percent of codes indicated each phase. For example, Kelly had 87.5% Expectation₁ phase during the house tour, which we calculated by taking the frequency count of her Expectation phase codes indicated during this session (HT) (7 instances) and dividing it by the total of phase codes in the session (8 instances total in HT) and then put it in percent form.

We defined characteristics of participants’ experience of acceptance phases in

Table 3. Defining occurrences of acceptance phases.

Occurrence	Definition
Primary Peak	The highest point of a given phase.
Secondary Peak	Any point that is higher than the points on either side and is not the primary peak. Must be 10% or greater. Points with no data on either side (like House Tour and Month 7) can be peaks).
Hill	Any point that is higher than the points on either side, but less than 10%.
Valley	Any point that is lower than the points on either side.
Ascent	Any point that is higher than the previous point.
Descent	Any point that is lower than the previous point.

—please review this table now. We defined participants as being in a given phase when that participant’s codes for that phase had a primary or secondary peak (Appendix D). If multiple phases had peaked at the same time, we described participants as being in multiple phases. We consider peaks as indicative of experiencing a phase, rather than which phase was most commonly discussed in the interview because some topics were more commonly discussed throughout the entire study. For example, the Identification₆ phase was the highest-indicated phase for most participants across several months; however, during those months, other phases peaked. If we ignored these peaks, it would seem as though participants did not enter these phases. There are several reasons that certain phases are more common than others. First, following the Phased Framework of Technology Acceptance (PFTA; [9]) coding scheme, the Identification₆ phase had eight codes whereas all other phases had between one and four codes [9], which made Identification₆ the most commonly indicated. Second, our interview questions may not have sufficiently encompassed questions about each phase.

Note that there were no assigned codes to the Encounter₂ phase in the PFTA [9], instead indicating that this phase always occurred during participants’ first encounter with the technology. Likewise, we assigned no Encounter₂ phase codes–but we did code the Bringing Robot interview, which was the participants’ first encounter with the robot. In our study, all participants experienced characteristics of the Expectations₁, Adoption₃, and Identification₆ phases during this session.

We also combined the PFTA’s [9] “intermediary” codes (i.e., Adoption/Adaptation_3·5, Adaptation/Integration_4·5) with the full phases (Adaptation₄ and Integration₅ respectively). These codes had very few occurrences, and doing so strengthened the appearances of these phases and more accurately represented the flow of phases, rather than having in-between phases that were underrepresented.

3.2. Participants Experience

Below we describe how participants experienced the acceptance phases as discussed in their interviews, month by month. We describe in depth any month that has any type of peak. All other months, or “transitional months”, we skip to save space. We discuss phases that were greater than 10% and Hills (<10%) that are the only occurrence of a phase for a given participant. Figure 2 and Figure A1 show the Phased Framework of Technology Acceptance coding totals across all four participants, as percentages and frequency count in 100% stacked column form, respectively;

Table 4. Results of participants’ primary peaks of phases. * Indicates hill of a phase when a primary peak did not occur.

Participant	Expectation	Encounter	Adoption	Adaptation	Integration	Identification	Non Use
Sasha	HT	BR	BR	M1	M6	M3	-
Kelly	HT	BR	BR	M1	M5	M4	M3 *
David	HT	BR	HT	M2	M5	M3	M1 *
Jessica	HT	BR	HT	M1	M6	M5	M3

shows interview corresponding to the peak of each phase per participant.

3.3. Sasha

Sasha was the youngest participant in the study (aged 63). She was married and had two daughters, one of whom lived with her and her husband while attending college. During the study, Sasha was still working. Her relationships were very important to her, and she had an active social life. She regularly met friends on the weekends and kept up with friends who have moved away by sending Tik Toks (Tik Tok—A social media platform, originated in China. A space where users create and post short video clips) and frequently texting. Her previous experiences with robots came from science fiction movies like Star Wars, and she expected herself to be leery of the mobile telepresence robot. Sasha used the robot with her interaction partner, Chasity_IP. Sasha progressed rather quickly through acceptance phases, with indications of Encounter₂, Adoption₃, Adaptation₄, and an ascent in Identification₆ all by the time she met the robot. After that, she fluctuated between phases, but her Identification₆ remained the highest (See Figure 3 &

Table 5. Sasha’s acceptance of technology phases.

Phase	House Tour	Bringing Robot	M1	M2	M3	M4	M5	M6	M7
Expectation1	PP	D	D	D	-	-	SP	-	SP
Adoption3	-	PP	D	-	-	-	-	-	-
Adaptation4	SP	-	PP	D	D	-	SP	-	SP
Integration5	-	-	-	A	A	SP	V	PP	D
Identification6	-	A	A	A	PP	D	D	D	SP
Non Use⌀	-	-	-	-	-	-	-	-	SP

).

3.3.1. House Tour Interview: Expectation₁

During the House Tour, as expected, Sasha showed her primary peak of the Expectation₁ phase (85.71%), with a secondary peak of Adaptation₄ (14.29%).

Sasha had positive associations with robots. She associated robots with “Google Assistant (Google Assistant—A virtual assistant software developed by Google. It is available on any mobile and home automation device)”, which she “use[d] … everyday … it’s helpful” (Association₁, Attitude Formation₁). Sasha also made associations with friendly robots from “the movies … [the robots are] part of the family … maybe one day it will be like that; many people are interested in that” (Association₁). She was familiar with some advanced technology that tells people “it is time for their medicine” (Association₁). Although Sasha had previously expressed feeling “leery” of the mobile telepresence robot, after meeting with the researchers for the House Tour, she felt that she would be comfortable with adjusting her home life to include it. She began preparing a clean open area that was safe from her cats (Preparation₁). Sasha also began developing ideas of how this robot would fit into her family life and encounters, like in the Adaptation₄ phase. Socially, Sasha enjoyed social gatherings every few weeks, but most of her close friends and family moved away over the years. Now, her relationships were primarily based around “… sending Tik Toks … on the phone and messaging each other”. Sasha thought that the robot could be another way to connect with her family: “I think that my grandchildren would love it … they are so inquisitive at [ages] five and three” (Novelty₄).

Thus, before Sasha had encountered the robot or knew what it would look like in her life, she had positive associations and attitudes about it and had the impression that she would accept it.

3.3.2. Bringing Robot Interview: Adoption₃

During the Bringing Robot interview, Sasha’s Expectation₁ phase (54.55%) began its descent and the Adoption₃ phase (18.18%) had its primary peak. The Identification₆ phase (27.27%) also ascended.

Sasha continued developing positive Expectations₁, which increased her Adoption₃ (i.e., Excitement₃). She was excited to see “all my kids [and], grandchildren”, who lived several hours away, as well as her daughter who had recently had a baby. She was “look[ing] forward to finding out how it will be to interact and cook together” through the mobile telepresence robot (Anticipation₁, Excitement₃). She was concerned about if her much-older in-laws could operate and understand this robot; they had had a hard time visiting with them since the COVID-19 pandemic (COVID-19—An infectious disease—discovered in 2019 and became a global pandemic 2020-2021. Many infected with the virus experienced a variety of symptoms. Some experienced worse symptoms leading to serious illness or death. Therefore, many governments recommended or mandated social isolation to reduce the spread of the virus) began (Anticipation₁).

Sasha also had to adjust her expectations of this robot, indicating Adoption₃, due to some slight connectivity issues. She had “thought it would connect smoothly …”, but recognized that like most other technology, “there are always issues when you get something new and try to get it to work” (Adjustment₃). Despite the issues, Sasha was highly enthusiastic about having the robot in her home and open-minded about the possibilities it brought into her life.

During this first experience with the robot, she showed early signs of Identification₆, by recognizing some benefits and disadvantages of the robot. She was concerned that her guests could “call me at a time when I don’t want” (Recognize Disadvantages₆). Still, she was enthusiastic that using the robot is “totally different because you don’t have to hold the phone; you can use both hands, and they [your friends or family] can talk with you and play games and stuff” (Recognize Benefits₆).

3.3.3. Month 1 Interview: Adaptation₄

In Month 1 of use, Sasha’s Expectation₁ (31.58%) and Adoption₃ (10.53%) descended. The Adaptation₄ phase rose to its primary peak (21.05%), and her Identification₆ (31.58%) continued its ascent.

During Month 1, Sasha’s Expectations₁ and Attitudes₁ regarding the mobile telepresence robot stabilized as she became more acquainted with the technology. One new attitude was that using the robot to communicate with Chasity_IP felt like they were “talking in-person”, and it was “nice to have company” via the robot (Attitude Formation₁). After this Month, most indications of Expectation₁ related to “discuss[ing] it [the robot] with others” and sharing her experience (Discuss with Others₁).

As Sasha used the robot, she needed to Adapt₄ to it. Such advanced technology was new to her, and this Novelty₄ led to some challenges with technical issues and fear of the robot. Sasha had “difficulty with the controls”, driving the robot, and connectivity issues like “disabled calls” (Trial and Error₄). These technical issues caused Sasha and her daughter some “frustrat[ion] because she could not swerve” (Trial and Error₄) while driving the robot. Due to the novelty, she also felt that “it was scary when it moved” (Novelty₄). Despite the issues in her first month, Sasha was dedicated to using this robot to improve communication with her family; Sasha wanted to strengthen her relationship with Chasity_IP (her daughter) and develop a deeper relationship with her grandchildren.

3.3.4. Month 3 Interview: Identification₆

Sasha had a primary peak of the Identification₆ phase (64.71%) in Month 3, while the Adaptation₄ phase (11.76%) descended and the Integration₅ phase (17.65%) ascended.

After 3 Months of use, Sasha was more comfortable with the mobile telepresence robot and came to Identify₆ with it. She expressed its Benefits₆ and Disadvantages₆. For example, one disadvantage was that “it cannot go downstairs” (Recognize Disadvantages₆); however, a benefit was that it “was not very bulky, [and] very lightweight”, which allowed her to carry it up and down her stairs (Recognize Benefits₆). Although she wanted to increase her use and “get a lot of people to know about” the robot, Sasha indicated that a major disadvantage was, “if they [other users] ask me how to do something, I can’t tell them” (Promotion to Others₆, Maintenance₆, Recognize Disadvantages₆) due to the limited instructions or user manual from the company.

While Sasha recognized some disadvantages, she overall felt that the robot was a positive part of her social life because it helped her to “see her grandchildren and interact with them”. When “Harold [grandson], who is five, discovered it, he really liked it” (Confirmation₆, Promotion to Others₆, Maintenance₆). Because of these positive experiences, Sasha concluded, “I don’t think I have anything negative on it” (Confirmation₆). In later months, she and the grandchildren also enjoyed “dancing together” (Exploration₄) through the robot.

3.3.5. Month 6 Interview: Integration₅

During Month 6, the Identification₆ phase, which had been descending since Month 3, was still high (47.06%). Sasha had her primary peak of the Integration₅ phase (47.06%).

By Month 6, Sasha had become more comfortable using the mobile telepresence robot in her daily life, using it regularly and incorporating it into newly established routines with her grandchildren, like using the robot to “play hide and seek … which was a lot of fun …” (Use Routines₅, Reinvention₅). She only wished that the robot could “call 911” if potential older users, like her in-laws, fall (Reinvention₅).

3.3.6. Final Notes

Sasha entered this study with no prior knowledge or experience with mobile telepresence robots, but with the positive mindset that “it was interesting” and the determination to “just go with it”. In her Final Interview, Sasha explained how her initial discomfort and challenges with the robot transformed into comfort and even Emotional Attachment₆ to it.

Throughout the study, Sasha adjusted her opinions about technology. She began with “trouble with sending the visitor pass or connecting … it needs to be more user friendly”, especially for “older adults who are older than me” (Adjustment₃, Curiosity₃). “When [she] first started using it I [Sasha] couldn’t figure out how to use it” but then her son-in-law came to her home to assist her in learning the robots controls (Trial and Error₄). With her son-in-law’s help, she figured out the controls and learned that his “Apple phone worked a lot better” with the robot than her Android phone (Personalization₄).

Sasha began the study “a little apprehensive, because [she] was not sure what the robot would do”. During the Final Interview, Sasha opened up about how, as she became more comfortable with the robot, she became increasingly emotionally attached to it, and how it affected her family. In Month 1, Sasha predicted “becom[ing] attached” and that the robot would “spoil” her but did not mention it in other monthly interviews. During the final interview, Sasha shared that she had gendered the robot as a ‘he’ and named it “Bob” (Emotional Attachment₆). She felt that “he [the robot] became part of the family … instead of being a piece of furniture, he became part of the circle … as you interact with family through him [the robot]”. She “became attached and I [Sasha] was sad when Daniel [researcher] was taking him [the robot] apart to take him [the robot] away”. Sasha was proud of her attachment with her robot, but “thought it was kind of strange to be able to be so close to it … something that’s not human … but it feels like it’s human”. She credited this emotional attachment to the robot “helping build the connection with my daughter”, and she found that her husband got “a little jealous” about her connection with the robot. Sasha appreciated the benefits of being able to “interact and see your family, especially if you are bedridden or you can’t get out of your home … that does wonders for your psyche not to be so depressed or so alone” (Recognize Benefits₆).

Overall, Sasha had a “positive experience” and reported that she would continue having this robot or one like it within her home if given the opportunity (Curiosity₃, Confirmation₆).

3.4. Kelly

Kelly (aged 70) was a very proactive participant; she actively sought ways to push her mobile telepresence robot to the limits and discover the possibilities of its use in her home and daily life. Kelly was married, lived with her husband and was retired. She spent much of her free time playing board games and video games with her friends. Of our participants, Kelly was most savvy about current technology and actively used services like online message groups and online gaming software. She kept in touch with several close friends via Discord (Discord—A chat app used to communicate with others via text, video, or voice) chats or in weekly meetups. She also socialized via dinners, catch ups, game nights, online gaming, and traveling to see loved ones in person. Kelly found some robots to be quite unique and purposeful and had knowledge of them from science fiction. She indicated no preconceived opinions of the robot and that she wanted to go through the study with an open mind. Kelly used the mobile telepresence robot with her interaction partner, Selena_IP. Kelly advanced through the acceptance phases rather quickly and experienced the Identification₆ phase within her first encounter with the robot; this phase was her highest ranked phase throughout the rest of her seven-month experience (Figure 4 &

Table 6. Kelly’s acceptance of technology phases.

Phase	House Tour	Bringing Robot	M1	M2	M3	M4	M5	M6	M7
Expectation1	PP	D	D	-	D	SP	-	-	SP
Adoption3	A	PP	-	-	-	-	-	-	-
Adaptation4	-	-	PP	-	D	-	-	SP	-
Integration5	-	-	A	-	SP	V	PP	D	D
Identification6	-	SP	V	-	A	PP	V	A	SP
Non Use⌀	-	-	-	-	H	-	-	-	-

; Month 2 is excluded due to missing data for that session).

3.4.1. House Tour Interview: Expectation₁

During the House Tour, Kelly showed her primary peak of the Expectation₁ phase (87.50%). The Adoption₃ phase (12.50%) was ascending.

Kelly’s Expectations₁ about the robot came from indirect knowledge. She had no real-world experience with robots, but had seen “them in movies and books, such as Star Wars or the Locke series by John Scouts”. She knew that “TV robots can be self-aware such as in Star Wars”, (Association₁), but also understood that real-life robots were different. Her knowledge of different types of robots included that some are “a mechanical self, or [some] you self-controlled … [they] can be programmed to do various things …” (Attitude Formation₁).

Based on her Expectations₃, she was positive about Adopting₃ the mobile telepresence robot. She thought it would complement her social life “because I can contact people … in real time … and not just on the phone or [by] email” (Anticipation₁, Excitement₃), which made her eager to begin Adopting₃ (12.50%) the robot.

3.4.2. Bringing Robot Interview: Adoption₃

During the Bringing Robot interview, Kelly had her primary peak of the Adoption₃ phase (30%) and a secondary peak of the Identification₆ phase (40%). Her Expectation₁ phase (30%) was steeply descending.

When Kelly met the mobile telepresence robot, she instantly began adjusting her previous Expectations₁. She recognized Benefits₆ and Disadvantages₆ of the robot, thus presenting a secondary peak of the Identification₆ phase (40%). She found that the robot “cannot back into the charging station”, which may be problematic if IPs were periodically checking in on older adults who may live alone or be homebound. She also wished that the tablet “could look up or down at something” (Recognize Disadvantages₆). However, she felt that these Disadvantages₆ were “not that big of a deal” in comparison to the Benefits₆: “It gives me better access to people who don’t live near me, … and it seems more interactive than either Zoom or … the phone” (Recognize Benefits₆). She also could not “think of any reason not to have it”. For Kelly, the Benefits₆ outweighed the Disadvantages₆; she showed Adoption₃ signals: “I am very excited about it, and I know that Selena_IP would be very excited by it as well. It’s going to be a lot of fun to hang out with someone” (Excitement₃).

3.4.3. Month 1 Interview: Adaptation₄

During Month 1, Kelly had her primary peak of the Adaptation₄ Phase (19.05%). The Expectation₁ phase (19.05%) was descending, and the Identification₆ phase (28.57%) was a valley, and the Integration₅ phase (23.81%) was ascending.

Having experienced positive Expectations₁ and Adoption₃ (Excitement₃) in the previous sessions, Kelly began Adapting₄ to the mobile telepresence robot during her first month with it, through Exploring₄ its various abilities and applications. This robot altered the way Kelly communicated with friends or family who lived in different states or countries, and she enjoyed that “it’s mobile, whereas with Zoom you just have to sit there …” The robot “can follow me around or I can follow you [the pilot] around” (Novelty₄). Kelly was more adventurous than the other participants. Although some days, Kelly and her interaction partner simply “hung out together … [without doing] anything extensive”, her primary goal was to “experiment to find new activities to do with” the robot and to “learn more possibilities for it” (Exploration₄). She often tried to discover the bounds of the robot’s capabilities like when she “tried [to take the robot] out for a walk [outdoors]” (Exploration₄).

Some of Kelly’s adventures worked better than others. She was still learning the operations and possible uses for the robot because it was such a new device to her. For example, Kelly expressed the early frustration of navigating it: “The most frustrating thing is getting it to move where you want it and when you want it” (Familiarization₅). She wanted to “learn the controls better for Selena_IP. She learned to drive the robot herself and become “more comfortable” with it, thus showing early signs of Integration₅. She “set up an obstacle course to get Selena_IP more comfortable with navigating it” (Familiarization₅). The valley of the Identification₆ phase at this time was because Kelly talked less about Benefits₆ and Disadvantages₆ of the robot compared to the previous; she Maintained₆ consistent use.

3.4.4. Month 3 Interview: Non Use_⌀ (Hill)

During Month 3, Kelly had a slight decrease in her usage, resulting in a Hill of the Non Use_⌀ phase (5%). Her Integration₅ phase (30%) had a secondary peak, and she still showed Expectation₁ (15%) and Adaptation₄ (10%) descending and Identification₆ (35%) ascending.

During this interview, Kelly reported decreasing use of the mobile telepresence robot: “I tried to, but I have been busy” (Lack of Use_⌀). This was Kelly’s only indication of the Non Use_⌀ phase. With this instance being such a low indicator, we believe that Kelly never truly entered this phase, as she used it more consistently after this interview.

During this interview, Kelly realized that she had developed a Use Routine₅ for using the robot (Integration₅), which she had fallen out of the habit of in Month 3 because she was busy. After this realization, Kelly strove to preserve her use and re-establish the routines she previously set in place. To better maintain these routines, she sought out new uses and further learned the controls for herself, discovering “the screen’s dark feature” (i.e., the night vision mode; Familiarization₅).

3.4.5. Month 4 Interview: Identification₆

In Month 4, Kelly resumed her normal Use Routines₅ and had her primary peak of the Identification₆ phase (58.33%). She had a small secondary peak of the Expectation₁ phase (16.67%) appear, much later than expected. She also showed a valley of Integration₅ (25%).

After Adapting₄ to and Integrating₅ the mobile telepresence robot into routines in the prior months, Kelly showed her primary peak of Identification6. This peak occurred based on her recognition of the robots Benefits₆ and Disadvantages₆ and was supported by the increase of the Expectation₁ phase (16.67%). Benefits₆ Kelly mentioned included, “it’s a nice size, it doesn’t take a lot of room, and … [its] application it’s pretty good” (Recognize Benefits₆). She also recognized some Disadvantages₆ of the robot that she considered “middle range” (i.e., to have a minor impact on her experience), such as “it is not very good outside”, and “it could include some application[s] that would make it more convenient” (Recognize Disadvantages₆). Understanding of the benefits and disadvantages also informed Kelly’s Expectations₁ of the robot: she did not “have any real negative” experiences as “it [robot] doesn’t come on in the middle of the night and roll around”(Attitude Formation₁), and she felt her negative experiences could not compare to something like that occurring.

Kelly also began Identifying₆ Selena_IP with the robot when it became their primary form of communication: It is “nice to hang out with Selena_IP and craft or play board games…. It’s a memorable or awesome thing being able to set up a board game or cook with a larger group of friends who have not seen her [Selena_IP] in a while” (Identification₆). This positive experience led to the second peak of Expectation₁; Kelly wanted to “find some more people to” use the robot and began looking towards her social circles for other interaction partners (Discuss with Others₁).

3.4.6. Month 5 Interview: Integration₅

Kelly’s Integration₅ phase (38.46%) reached its primary peak. Identification₆, still her highest ranked phase, entered a small valley (46.15%).

With her primary peak of the Integration₅ phase (38.46%) in Month 5, Kelly was continuing her use of the mobile telepresence robot at least “once a week” (Use Routines₅) if not more than in Month 3 where the first peak of Integration₅ appeared. Now, Kelly and her interaction partner wanted to “play board games over this [the robot] … and do something new” (Exploration₄). Following her Reinventive₅ mindset, Kelly constantly pushed the bounds of what the robot could handle. She learned, and taught others, who she had previously introduced to the robot, how to “play board games … and teach others how to use it [the robot] and have some fun” (Reinvention₅). She wanted this robot “to connect to a hot spot” for a more on-the-go experience, attempting again to go on a walk outdoors with her robot (Reinvention₅). Kelly continued this Reinventive₅ mindset through the rest of the study, searching for new and exciting ways to use this robot like a board game night where it “feels like having an actual game night” (Use Routines₅).

3.4.7. Final Notes

Kelly entered this study with an interest in robots and research due to her previous experience: “I had taken part in another study regarding robots … that made me interested when I heard talking about robots and research; I want to be part of something like that” (Curiosity₃). By participating in the study, she hoped to connect with her friend Selena_IP, who moved away, and the study met her expectations. Kelly felt that she had “limited myself with the study by mostly doing things just with Selena_IP”, but felt that this mobile telepresence robot could allow her to “make a lot more connections with other people as well” (Recognize Benefits₆) if she had dedicated more time to working with others outside of her selected IP.

Kelly used the robot “about once per week” (Incorporation₅) for various activities: “Sometimes we played games and sometimes we crafted. Sometimes we would just visit and chat” (Reinvention₅). Her most positive experience was “being able to get together with Selena_IP and [others] and play board games. We all live thousands of miles apart, so it was nice” (Use Routines₅). What made the experience so positive was “just being close with them and visiting them, because we are friendly” (Identification₆). Another positive experience was teaching Selena_IP how to make sushi via the robot: “The overall experience of making sushi with Selena_IP was great. Selena_IP was rolling around the kitchen looking at what we were making and our tools” (Incorporation₅). Kelly “did not have any negative experiences other than a couple of glitches”, like when it was “hard to hear”, and parking issues. She offered some improvements for the robot like a “better base”, “bigger screen”, and “voice control” (Recognize Disadvantages₆).

Although Kelly connected with her friend through the robot, she never bonded with the robot itself. She had a strong opinion that she would not “want to form a bond with a non-living thing”, adding that the robot “is a great way to contact those who are not nearby, but that is the end” (Confirmation₆). Towards the end of the study, Kelly Discussed₁ the robot with others and received “mixed reactions” as some “thought it was really cool and interesting … [and] others had reservations about the robot” (Discuss with Others₁).

3.5. David

David (widower), lived alone and was our eldest participant (aged 84). He was our only participant with past experience with robots, and he had a strong passion for robotics and technology. In the 1980s, working as a communication technician for a company, the Federal Aviation Administration (FAA) in Alaska, he assembled robot components and tracked communication with the pilots. During that time, he helped implement robots in high schools. Early in his career, he worked as a sales manager of cell phones and pagers as their popularity rose. After retirement, David regularly met past colleagues and friends who had similar backgrounds or worked in related fields, like Engineering. He enjoyed frequently sharing and discussing his robot with them. David used the mobile telepresence robot with his two interaction partners, Michelle_IP and Patricia_IP. David was also very passionate in his interviews about his beliefs that “a lot of [older adults have] fear and confusion [about technology, which] must stem from ‘computer illiteracy’” and people getting “nervous about privacy”. Despite some issues with his robot’s battery in Month 1, David had a positive experience and moved very quickly into later acceptance phases. He used the robot with two different interaction partners (Figure 5 &

Table 7. David’s acceptance of technology phases.

Phase	House Tour	Bringing Robot	M1	M2	M3	M4	M5	M6	M7
Expectation1	PP	D	D	D	-	-	SP	-	SP
Adoption3	PP	-	SP	-	-	-	-	-	-
Adaptation4	-	-	-	PP	-	-	-	-	-
Integration5	-	SP	V	SP	V	A	PP	V	SP
Identification6	-	A	SP	A	PP	D	V	SP	D
Non Use⌀	-	-	H	-	-	-	-	-	-

).

3.5.1. House Tour Interview: Expectation₁ and Adoption₃

During his House Tour, David had primary peaks of Expectation₁ (71.43%) and Adoption₃ phases (28.57%).

David based many of his Expectations₁ of this mobile telepresence robot on his prior experiences working with robots for the FAA. He felt there was “a lot of confusion about the word robot …” He described a robot as “typically something that a human operator remote [controlled] … an autonomous robot is not really a robot; it’s an autonomous device that does its own thing” (Association₁). David looked forward to his encounter with the robot and “see[ing] what it is capable of” (Anticipation₁).

These Anticipations₁ directly led to Adoption₃ as his Excitement₃ surfaced. This “technology is interesting”, he said, adding that he will “just have to see it to know” if the robot could compliment his social life (Curiosity₃, Excitement₃). Having past positive experiences with robots, David sought more information about it and embraced positive feelings about the robot’s potential. He used this House Tour to prepare himself and his home for the robot, adjusting preconceived ideas of floor space needed and the movement abilities of this robot (Adjustment₃). Later, during David’s first encounter with the robot (Bringing the Robot), he noticed that the “camera seems to have the ability to look up …” (Exploration₅).

3.5.2. Month 1 Interview: Non Use_⌀ (Hill)

In Month 1 of the mobile telepresence robots’ use, David’s Expectation₁ phase (26.67%) descended and Integration₅ phase (13.33%) had a valley, while secondary peaks occurred for the Adoption₃ (20%) and Identification₆ phases (26.67%). Concurrently, David showed a hill of the Non Use_⌀ phase (6.67%).

David had a positive and informative introduction to the robot in the Bringing Robot interview and immediately began using it regularly. David had a secondary peak of the Adoption₃ phase (20%) as he dove into exploring and sharing the robot with his loved ones. A family down the road from him with older children had previously “sort of adopted” him. He introduced the robot to them, and they did “silly stuff” like playing chase and hide and seek with the children (Adjustment₃, Excitement₃, Exploration₃). In doing so, David Adjusted₃ his expectations and learned what the robot was capable of for his future use.

Unfortunately, the robot developed battery issues, preventing David from using it the last few weeks of that month. While he never truly entered the Non Use_⌀ phase, during Month 1 these technical issues caused a hill (6.67%) while he did not use the robot (Lack of Use_⌀). We quickly replaced his robot, resolving the battery issue, and he returned to using the robot regularly.

3.5.3. Month 2 Interview: Adaptation₄

During Month 2, David had his primary peak of Adaptation₄ (20%). He had a secondary peak of Integration₅ (30%). Meanwhile, Identification₆ (30%) ascended and Expectation₁ (20%) descended.

Because of his Month 1 battery issues, David’s experience with and acceptance of the mobile telepresence robot were slightly delayed compared to Sasha and Kelly, who experienced Adaptation₄ in Month 1. However, once he received the new robot, David used it to its full capabilities and had his primary indications of the Adaptation₄ phase. He Explored₄ using the robot as he aimed to better understand how he could use it in his life and struggled to operate it within his home (Trial and Error₄). His biggest complaint was a lack of information or inclusion of a user manual by the company (Double Robotics): “They don’t give you much information … you have to try it out yourself, and that’s fine with me, but probably a deal killer for some people” (Trial and Error₄, Exploration₄, Recognize Disadvantages₆). David questioned the design of the robot and suggested for the company to alter the build to include a “drive system like a tractor” (Association₁). Ultimately, David expressed that he enjoyed his use of the robot (Attitude Formation₁), but he had more questions before deciding how to use it beyond Month 2. David also expressed a concern for safety and privacy if individuals might “turn it on in [inappropriate] situations” (Recognize Disadvantages₆).

3.5.4. Month 3 Interview: Identification₆

In Month 3, David’s Integration₅ phase (15.38%) had a valley and the Identification₆ phase had its primary peak (76.92%).

Because of David’s Exploration₄ and Trial and Error₄ during Month 2, in Month 3, he could better Identify₆ distinct Disadvantages₆ and Benefits₆, and the Identification₆ phase became his highest indicated phase for the rest of his experience. David, our only participant with prior real-world experience with robots, provided a lot feedback regarding the issues this robot faced when implemented into a home setting. He felt that this robot needed a better wheel design because “if it has to go over an extension cord … it is shaky … because of the wheels” and because it struggled to drive from rug to hard flooring (Recognize Disadvantages₆). He knew that this robot was officially designed for “running around office suites” but found it a drawback that “you need to run it [robot] on totally level floors … and you can’t go outdoors” (Recognize Disadvantages₆). He also had issues with the battery not holding its charge (Recognize Disadvantages₆). Although David expressed many disadvantages, he firmly believed that “the potential is certainly there” with this robot and that “it is quite highly developed”, though further design updates would greatly benefit users (Recognize Benefits₆, Confirmation₆). David added, “when it works, it’s great”, he had “never seen anything like it”, and that this robot is “quite valuable for people in my situation [older adults living alone] …. I am a old guy, and I can roll a gasket or something and be on the floor for two weeks before anybody knows” (Recognize Benefits₆). At this point of the study, David had fully Incorporated₅ the robot into his life and was using it “not daily but a couple times” a week (Incorporation₅).

3.5.5. Month 5 Interview: Integration₅

In Month 5, David’s Identification₆ phase dropped into a valley (50%). The Expectation₁ phase (14.29%) had a secondary peak and the Integration₅ phase (35.71%) had its primary peak.

Although the Identification₆ phase was the highest expressed phase in Month 5, it was in a valley. David spent five months Familiarizing₅ himself with the mobile telepresence robot, exploring its possible uses, and increasingly Incorporating₅ it into his routines. Prior to possessing the robot, David kept in touch with his close friends and family regularly. Now, his use of the robot in weekly routines was at an all-time high as it was a means of continuing this communication, using it at least “two times a week for chit chatting and visiting with friends” and family (Incorporation₅ and Use Routines₅). David used the robot to visit several different people, who “all move it and control it differently” (Use Routines₅). The robot became “an auxiliary member” of his household (Personality Attribution₆). During his family reunion, David showed Expectation₁ (14.29%) phase characteristics when he discussed his experiences with the robot with his whole extended family (Discuss with Others₁).

3.5.6. Final Notes

David entered this study with “great interest in robots” (Curiosity₃) and experience with “having one in the 80’s … in schools”. Of our participants, he was one of the most interested in the study purpose, hoping to “learn something about the use of robots in terms of isolated elders … This study is a worthwhile thing to explore” (Anticipation₁). Before the study, he had been “on all kinds of platforms [the Internet], and a lot of them mention the use of robots in bringing people together, and I wanted to check that out” (Anticipation₁).

Overall, David’s had a positive experience; he appreciated the advantages of connecting with people through the robot: “If you are sitting here with a laptop and you’re on Zoom, that’s wonderful technology itself, but it is somewhat limited. Using the robot gives me a better picture of at least my home environment, and we can all walk around, and you can send it to different rooms” (Recognize Benefits₆). David’s main negative experience was that “there was a battery failure. That was a little scary … [the robot] toppled over and hit the floor. But the company was familiar with the problem and took care of it” (Recognize Disadvantages₆).

David’s comfort with the telepresence robot increased over time because the robot “is well designed, and I don’t have any fear of technology, so it [comfort] would be what I would expect” (Confirmation₆). David would “give this robot 10 out of 10 [stars] for the specific purpose that it was designed for”—that is, to be used in office spaces or a classroom for virtual work or learning. David enjoyed the robot so much that he regularly used it several times per week with friends and family who lived in different states or just down the road. David enjoyed showing the robot to the neighborhood kids and letting them use it to communicate and play hide and seek.

Finally, David did not report any emotional connection to the mobile telepresence robot: “It’s just a piece of gear. It’s not an unwelcome addition to my household as it takes up very little room, but you don’t want to get too emotional about technology. It’s just a tool” but he found that “walking in the door and seeing someone on the screen” was nice (Recognize Benefits₆). David suggested that this robot should have different applications specifically for the “isolated elderly population”, like “an autonomous feature like a security system” that users could have on a “schedule to check the premises when nobody is home” (Reinvention₅). He further suggested the robot to have the ability to follow specific instructions such as “play music” or dance” (Recognize Disadvantages₆).

3.6. Jessica

Jessica (aged 75) lived with her husband and son. She was our only participant to indicate full Non Acceptance_⌀ of the mobile telepresence robot. Jessica was familiar with ‘gadgets’ like Instacart (Instacart—online delivery service platform that allows users to have items delivered to their homes) and smart clocks (Smart Clock—A wireless clock that can connect to the internet for information). This influenced her opinions on robots as she believed robots are “tools … that help and do stuff for you”. Jessica used the mobile telepresence robot with her interaction partner, Victoria_IP. Jessica was disappointed that the robot did not have more diverse applications, as she had expected it to assist her in her home with tasks like searching for recipes online while cooking. Throughout her experience with the robot, Jessica had trouble operating it on her own. This made her confused and agitated throughout the study. She explained this Lack of Use_⌀ with her belief that she was experiencing a study manipulation that limited her access to the robot and that she was in a different condition compared to Kelly, a participant she personally knew. Despite the researchers’ best efforts to show Jessica how to use the robot and explain that there were no study manipulations, she maintained her beliefs throughout the study. She only used the robot during monthly interview sessions when researchers helped her set up her interaction partner on the robot. She concluded that there were “no benefits at all from this robot” as no one could “access” it (Non Acceptance_⌀; Figure 6 &

Table 8. Jessica’s acceptance of technology phases.

Phase	House Tour	Bringing Robot	M1	M2	M3	M4	M5	M6	M7
Expectation1	PP	D	D	SP	-	SP	D	-	-
Adoption3	PP	-	-	SP	-	-	-	-	-
Adaptation4	-	-	PP	-	-	-	-	-	-
Integration5	-	-	-	-	-	-	-	PP	-
Identification6	-	A	SP	D	D	-	PP	V	SP
Non Use⌀	-	-	A	A	PP	D	V	SP	D

).

3.6.1. House Tour Interview: Expectation₁ and Adoption₃

During the House Tour, Jessica had primary peaks of both the Expectation₁ (85.71%) and Adoption₃ (14.29%) phases.

Jessica began this study viewing robots as “tools” that “have amazing capabilities … [that] are not always positive …” and that it is a matter “of how they’re programmed and human habit” (Attitude Formation₁). She was familiar with internet apps and robots from science fiction but knew that it was “obviously fiction”. She based her expectations of the robot on her Shark robotic vacuum (Shark robotic vacuum—A robotic vacuum cleaner that uses sensors to detect obstacles. This vacuum uses sensors to map floor plans and clean surfaces) (Association₁). Due to her limited knowledge of real-world robots, she expressed Curiosity₃ with this robot as she did not understand “what it does” or “how I would use it” (Curiosity₃).

3.6.2. Month 1 Interview: Adaptation₄

In Month 1, Jessica’s Expectations₁ (33.33%) descended as the Adaptation₄ phase (11.11%) had a primary peak, and the Identification₆ phase (44.44%) had a secondary peak. The Non Use_⌀ phase (11.11%) began ascending.

During Month 1, Jessica did not use the mobile telepresence robot outside the Monthly session (Lack of Use_⌀). She had trouble Adapting₄ to the robot because she was “very confused about the robot” due to “a lot of issues with the robot moving itself, not being able to work, [and the] battery” (Trial and Error₄). Jessica primarily discussed the lack of mobility, accessibility, and uses of the robot in her life. She also distrusted the robot’s structure. “… the robot would not stand on its own—it rocks—I don’t like that …”. She repeated that the “application is very limited … more limited than I thought it would be” (Recognize Disadvantages₆). The peaks of Adaptation₄ and Identification₆ did not indicate that Jessica was progressing to acceptance, but rather towards Non Use_⌀. In Month 2, Jessica had concluded that she was in a separate ‘condition’ of the study with “no access to the robot” to use on her own (Attitude Formation₁, Lack of Use_⌀). She often Discussed₁ with a fellow participant (Kelly) about the robot and the uses Kelly had found for it. Jessica formed new opinions about the robot: that it has “a design issue” and the only reason she would continue using this robot was for compensation (Attitude Formation₁).

3.6.3. Month 3: Non Use_⌀

In Month 3, Jessica’s Non Use_⌀ phase (83.33%) had its primary peak, and Identification₆ (26.67%) descended.

Jessica’s signs of Lack of Use_⌀ during the Bringing Robot interview peaked in Month 3. She continued not using the robot between interviews, believing she had “no access to it [robot], especially as the host” (Lack of Use_⌀). She was still disappointed that she could not use the robot for other things she had expected to do (e.g., move it herself). These expectations not being met resulted in Jessica’s Lack of Use_⌀, developing into Non Acceptance_⌀ and pushing her move deeper into the Non Use_⌀ phase. After a month of Lack of Use_⌀, Jessica moved into full Non Acceptance_⌀ of the Non Use_⌀ phase. She decided to only use it at monthly activity and interview sessions for her “payments” (participant compensation; Non Acceptance_⌀, Attitude Formation₁).

3.6.4. Month 5 Interview: Identification₆

During Month 5, Jessica’s Non Use_⌀ fell into a deep valley (20%) while the Identification₆ phase (55%) became a primary peak. The Expectation₁ phase (15%) descended from Month 4.

As Jessica’s Non Use_⌀ fell into its valley, her Identification₆ phase ascended, but it did not signify acceptance; she primarily Recognized Disadvantages₆ and Confirmed₆ her previous negative attitudes. Jessica described that the mobile telepresence robot was “frustrating” due to the “technological problems”, that “the controls are wonky”, and that the robot “is limited in terms of moving around, [and] it can’t park itself” (Recognize Disadvantages₆). Her difficulties connecting with Victoria_IP due to internet issues were “very frustrating” (Recognize Disadvantages₆). She repeated her previous attitude that the robot was “much dumber” than she had expected it to be (Confirmation₆); it was just “making a phone call”, and she wanted “the software to be easier to use and have more interfaces” (Recognize Disadvantages₆) and capabilities (Confirmation₆, Personality Attribution₆). Despite these Disadvantages₆, Jessica felt that using the robot is “kind of fun” (Attitude Formation₁), but she confirmed that she would only continue using it for the study’s monthly payments (Attitude Formation₁).

At this point, we had hoped we had cleared her confusion with thorough explanations of the robot and its controls during every monthly session and hoped that the next visit would be more positive, yet this interview only confirmed her previous attitudes and opinions of the robot and continued her towards a path of rejection and Non Use_⌀ (Confirmation₆, Lack of Use_⌀). Overall, Jessica’s negative experience was due to a lack of understanding of the technology and study overall, which sparked some frustrations with her entire process. These frustrations led to an increase in her Lack of Use_⌀ of the robot and ultimately set the rise of the Non Use_⌀ phase which resumed being her highest identified phase from Month 6–7. Despite her frustrations, Jessica continued using the robot during monthly sessions.

3.6.5. Month 6 Interview: Integration₅

In Month 6, Jessica’s Identification₆ phase fell into a valley (28.57%) and her Non Use_⌀ phase (57.14%) reclaimed the position as the current highest indicated phase as a secondary peak. She also had a primary peak of the Integration₅ phase (14.29%).

Jessica returned to her Non Use_⌀ in Month 6 and “totally ignores” the robot between sessions. She showed minor evidence for the Integration₅ phase in that she continued her monthly interviews, primarily for payment, and because she enjoyed the monthly shared activities of the “book club with Victoria_IP” (Use Routines₅). For Jessica, the only tangible benefit from the study was “getting to know Victoria_IP better” and “reading new books you otherwise would not have” (Use Routines₅), but she would have preferred to do this via videoconferencing. Her disappointment with the robot continued, as she could not “think of a use for it” in her life (Confirmation₆), concluding that the robot was “too limited” (Recognize Disadvantages₆, Non Acceptance_⌀).

3.6.6. Final Notes

Jessica was the only participant who continuously had a negative experience with the mobile telepresence robot and found little use for it. She started the study with “interest in what’s going on in the computer world” (Curiosity₃) and expected to “be using the robot in my household … [as] a way to access information conveniently. [However], it’s basically a tablet on wheels” (Attitude Formation₁). Starting in her first encounter with the robot, she was dissatisfied because it did not meet her Anticipations₁, and she felt it only had Disadvantages₆: “part of the point is to move the thing. I can’t move it. Victoria_IP moves it, so it’s a little confusing as to what the point is” (Recognize Disadvantages₆). She did not have “a particular function for it in my [Jessica’s] life … aside from that, I enjoy chatting with Victoria_IP” (Recognize Disadvantages₆). Jessica concluded that the robot was “taking up space and blocking an electric outlet that I was already using” (Recognize Disadvantages₆).

Throughout the study, she primarily used the robot for “book club” (Use Routines₅) and only during monthly shared activities. She was delighted to have gotten “to know Victoria_IP better” (Identification₆) and “read some books she had not otherwise would have read” (Excitement₃, Identification₆). However, she did not feel more present using the telepresence robot than video conferencing and found “no benefit at all from this robot” beyond what other technology offered (Non Acceptance_⌀). Her negative experience was because of her mismatched expectations of the robot’s capabilities (“it has no utility”; Non Acceptance_⌀) and her technical difficulties (e.g., when “the thing [robot] turned itself on and got off the stand”). Jessica expressed her delight to be finished with the study and have the robot removed from her as “no one had access to it” so that she could better use the space it was occupying (Non Acceptance_⌀).

The way that she felt about the robot as well as the usage had not changed throughout the study for Jessica. “I had no use for it, I couldn’t use it. [It was] just sitting there” (Non Acceptance_⌀). However, she did say that her interaction participant got more comfortable with the robot. The “first session was a bit awkward, [but] after that Victoria_IP picked up how to drive it”. Curiously, Jessica’s interaction partner named the robot after Jessica’s mother’s name. However, when asked if an emotional connection with the robot was or could be formed, Jessica answered: “In what universe? Not under the current conditions.” (Non Acceptance_⌀).

4. Discussion

In this case study, we examined how four older adult participants progressed through technology acceptance during a seven-month long study, more deeply analyzing data from prior work [10]. Results showed that three of the four participants accepted and integrated the technology into their lives during this study, while one participant rejected the technology and stopped using it. In the below sections, we discuss what hindered and what advanced technology acceptance for these older adults (RQ1). We identify key personas that we saw in this study (RQ2). We also discuss how we chose to identify when phases were occurring and how to identify which phase was the primary phase in an occurrence when multiple phases occurred at once (RQ3).

4.1. RQ1: Progression Through Acceptance Phases

Below we discuss important aspects of our participants’ contexts that influenced their progression through the acceptance phases. While prior work includes many models of technology acceptance (e.g., Technology Acceptance Model (TAM; [41]), Extended TAM (ETAM; [42,44,74]), Unified Theory of Acceptance and Use of Technology (UTAUT; [45]), Theory of Reasoned Action (TRA; [36]), Theory of Planned Behavior (TPB; [37,38]), Social Cognitive Theory (SCT; [39])), they tend to examine acceptance over the short term. In this paper, we contribute to the literature by indicating the main factors of longitudinal technology acceptance that emerged over seven months using a bottom-up approach. These factors did not fit neatly into any model of technology acceptance—although some had similarities to factors from various models, which we discuss below. Our factors were as follows: solving problems with technology, life situations, and personality/mindset. We recommend that future researchers consider these factors in longitudinal technology acceptance models.

4.1.1. Solving Problems with Technology

How people experience and solve problems with technology affects their acceptance of it. This begins with users’ hands-on encounters with the technology based on issues they have with it and its system characteristics (e.g., relative advantage, complexity; TAM). Positive encounters (including successful experiences and support with learning the technology) advance acceptance, whereas negative encounters (e.g., technological failure, especially that users cannot overcome) hinder acceptance and can ultimately lead to discontinued use [75,76,77]. Thus, even if there is a technical failure, if the users can overcome it, it can become a positive encounter. Several factors help people overcome technical issues: Prior experience with technology, (distinct from the TAM’s experience, which refers to prior experience with the current technological system), Computer literacy (ETAM; [42,44])—how well individuals generally understand technology-, Availability of resources (TPB)—including accessible instructions and people to help, and self-efficacy (SCT) or people’s confidence in themselves for solving problems. These overall led to perceived ease of use (TAM; [77]). In our study, while technical issues could slow acceptance, overcoming these issues (using factors described above) led participants to feel higher self-efficacy and ease of use, helping them advance toward full acceptance.

What hinders advancement? Participants faced issues with the mobile telepresence robot that hindered acceptance in two primary arenas: technical issues (i.e., unstable Internet connections due to coverage and/or bandwidth) and physical issues with the robot itself (i.e., lack of user manual, battery issues, and limitations of the device). David’s robot’s battery failed in Month 1, leaving him unable to use it for two weeks. This delayed his exploration of the robot and progression through the acceptance phases, compared to Sasha and Kelly. Jessica faced technical difficulties like not being able to sign onto or operate the robot. When such issues arose, Jessica’s limited prior experience with technology impeded working through this. Prior experience with technology relates to being able to use outside resources to solve technical problems because of needing to do this with past technologies [66,75,76,78,79,80,81]. Likewise, lacking social resources (i.e., people who could help) further hindered participants. Jessica had minimal knowledge and experience of advanced technology beyond fictional movies and her simple robotic vacuum. Although she lived with her husband and son, neither helped her understand or operate the robot. With these negative factors in play, she experienced low self-efficacy and perceived ease of use and stopped trying to fix technical problems. Because Jessica could not use the controls and basic functions of the robot, she never fully experienced the technology and accepted it.

What aids advancement? Even if system issues arose, if participants took the initiative to improve their understanding and comfort with the robot outside of working with the research team (self-efficacy and using resources), their increased knowledge helped them advance through the acceptance phases. This could be seen as creating a positive cycle of self-efficacy (SCT; [39]), as users have positive reinforcement for improving the system themselves, or with the help of those around them. Users could better do this when they had more prior experience with technology. Which ultimately aided users gain comfort with new technology. Participants’ prior experience with advanced technology (e.g., Kelly’s online gaming and Discord (Discord—a voice, video, and chat application for communicating and building communities across the world—publicly based common goals, hobbies, classes, or privately for friends or family)) and David’s work with social robots and frequent discussion of new technology) helped them understand the robot and navigate its limitations. They were also increasingly eager to learn more about the robot and test its capabilities, which gained them further experience, supporting their understanding and use of it. Although David had technical (battery) issues early on, he benefited greatly from using external sources to broaden his understanding of the robot and its operations, like YouTube (YouTube—video-sharing website for users to create, upload, and comment on videos) videos of the robot and communication with the Double Robotics team. His research on the robot helped him incorporate it into his routines. Kelly and David both dedicated time to learning the robot’s controls and limitations to support their interaction partners’ (IPs’) use. In Month 1, Kelly set up an obstacle course for her IP to practice navigating the robot. David practiced driving the robot so he could better explain it to his IPs. Although Sasha, like Jessica, had minimal knowledge and experience of advanced technology beyond fiction—only Google Home Assistant (Google Assistant—virtual assistant software developed by Google. It is available on any mobile and home automation device)—she had social resources to help. Sasha’s son-in-law visited several times to help her learn to use the robot. Through this support, Sasha gained self-efficacy about operating the robot. The knowledge that David, Kelly, and Sasha gained from resources outside of the research team helped them feel more ease of use operating the robot and with it in their homes and aided their advancement through the acceptance phases.

4.1.2. Life Situations

Various life situations, especially related to how much time people have to devote to learning and using technology, can hinder or aid technology acceptance. Such life situations from prior work include age [45,82], gender (UTAUT; [45,83]), marital status [84], employment status (employed/retired or part/full-time; [85]), and economic condition related to family support [84]. Also, habit (from [86], an extension of the TRA and TPB)—such as forming habits of using the technology—can influence and predict behavior [87]. In our study, we found that being busy, or having many social obligations related to less use of the technology—but that habits or social routines with it related to increased use.

What hinders advancement? Life situations—like work, frequency of social events, and general business—hindered the use of the mobile telepresence robot for social purposes due to having less time to dedicate to the technology and explore its uses. Because Sasha was working full-time, it was difficult for her to fit robot use into her routine. Although Kelly was retired, her filled social calendar sometimes limited robot use. Limited availability slowed these users’ acceptance as they both experienced brief periods of decreased or Lack of Use₇ due to “being busy” (Sasha and Kelly). In contrast, David was retired with a less active social calendar and dedicated more time to the robot, using it several times per week throughout the study. This pattern of business corresponding to reduced robot usage may relate to the robot’s use in mobile communication. If it further benefited busy individuals, like by cooking and cleaning, busy users may use it more often.

What aids advancement? When participants’ mobile telepresence robot usage decreased, scheduled routines and appointments helped them get back on track with using it, including using the robot for the study session or scheduled routines with family. Kelly returned to her robot use routines when her social life slowed down. Similarly, Sasha returned to her routine robot use to meet with her grandchildren, who enjoyed playing with the robot. David maintained regular routines with his IPs (both relatives) and friends in his neighborhood. Even Jessica, who disliked the robot, returned to its use for routine study sessions when the researchers could help her set it up (mostly to receive study payments). Established routines with loved ones and the technology gave users designated time with the robots, thus making them robust to disuse. Therefore, even though Sasha, Kelly, and David each had a short period of Lack of Use₇, it did not deeply hinder their acceptance.

4.1.3. Personality/Mindset

An individual’s mindset when trying something new can set the path for their acceptance or rejection. Although we did not measure personality in the study, some of our participants’ experiences relate to the Big Five personality categories, including agreeableness, openness to experience, extroversion, and neuroticism [88], which affect technology acceptance [78,89,90]. These relate to how people create expectations of technology and respond when expectations do not match reality [91,92]. Similarly, social influence like peer pressure can influence people to use the technology (UTAUT; [45]). In our case study, participants’ high agreeableness, openness to experience, extroversion, and low neuroticism aided advancement, while the reverse hindered it, as in past work [90]. In particular, we found that openness to the technology allowed participants to experience it without strong expectations, which advanced acceptance.

What hinders advancement? How people form and use expectations can hinder their advancement. Some users do not advance through phases due to a gap between their expectations of the technology and its actual capabilities [61]. Jessica had high expectations of the mobile telepresence robot, which may come from her limited experience with advanced technology [66,75,76,78,79,80,81] and importing expectations from fictional robots. She (falsely) expected that this robot would like a smart device that she could direct and struggled to understand its use and applications. The research team used multiple attempts and methods to explain the robot’s abilities and how to use it. However, Jessica’s high neuroticism (or pessimistic attitude) and low openness to experience hindered her acceptance hindered her interest in learning. She came to believe that it was not possible for her to use the robot and enter the Non Use_⌀ phase. Some of Jessica’s struggles with acceptance also relate to her active attempts to seek social approval from Kelly, whom she knew outside the study. Jessica talked to Kelly about the robot and often made frustrated comments to the researchers about not understanding how Kelly “had control” of the robot when she herself had no access to it. Jessica appeared to crave the experience that Kelly was having, but because she could not, she settled on an understanding that her access to the robot was purposefully blocked.

What aids advancement? A positive and open mindset and high levels of agreeableness can allow users to accept the limitations of the technology and seek applications that push the bounds of its limitations. Sasha’s and Kelly’s openness to experience and agreeableness helped them advance through the acceptance phases and gave them resilience when there was a problem with the mobile telepresence robot. Sasha, like Jessica, had little experience with new technology. She began her seven months with some fears related to this novel robot technology and challenges she began facing with using it, like control and internet connectivity issues. However, Sasha remained open to this new experience, which allowed her to resolve such issues and advance through the acceptance phases. Further, through her agreeableness, Sasha formed an emotional attachment to the robot and identified it with her daughter (IP). This helped her push through issues with the technology, because she enjoyed using it with her IP, and because it became “part of the family”. Kelly was open to experiences and upbeat. She actively sought new ways to use the robot and made a game of learning and pushing its limitations. Her routine use and imagination sped her progress through the acceptance phases by allowing her to use the robot in her daily life in a variety of ways.

4.2. RQ2: Participant Emergent Personas

Because people’s different characteristics and life experiences affect technology acceptance, as demonstrated above, it is important to consider a variety of people when designing technology. Creating and referring to user personas—i.e., models of potential users’ identities—can help researchers and practitioners understand potential users and develop technology for individuals’ specific needs and preferences [93,94,95].

Based on our detailed analysis of participants’ experiences across seven months, we developed emergent personas to provide an outlook for individuals’ experiences through the acceptance process and to predict which users may reject the technology. To identify these personas, we first summarize their general “attitudes”. Next, we draw on the user characteristics identified above: solving problems with technology (technology experience, support from others), life situations, and personality/mindset (personality, and emotional connection). We indicate how these relate to progress through the acceptance phases. We propose three personas in this paper: Sally the Standard, Addie the Advanced User, and Nora the Non-Starter. While we do not suggest these become industry-standard personas, we offer them as examples based on our participants. We recommend that future researchers expand on this with a larger sample, including more demographic variations, to illustrate a more comprehensive set of personas.

4.2.1. Sally the Standard

• Attitude: I am open to new technologies. I may be a bit worried about it, but I understand and accept that I will make some mistakes and the technology will have some errors, especially when I first start setting it up.
• Technology experience: Moderate: I am familiar with some basic and advanced technologies, like social media and Google Assistant. My familiarity with other devices can help me learn new technologies.
• Support from others: My family members or close friends help me learn new technologies—Especially when I am first starting or when new features are released.
• Life situations: I like to keep a busy social calendar. Sometimes it’s a little overwhelming.
• Personality: I am agreeable and open to new experiences like learning to use new technology. I am willing to look past some technical errors. I am not too adventurous and tend to stick with the typical or recommended uses of technology rather than come up with new uses for it.
• Emotional connection: I really appreciate the technology that helps me. After becoming familiar with it, I may find myself emotionally connected to it—especially if it is very useful or connects me to people I love.
• Progress through acceptance phases: I steadily move through the acceptance phases. Sometimes I stall when I get busy, but if I have set up use routines, they will draw me back to using it and help me achieve final acceptance.

4.2.2. Addie the Advanced User

• Attitude: I am very excited about new technologies and love being the first person to try them out. I want to discover all the possible use cases, and I make a game of finding new uses and limitations of the technology.
• Technology experience: High; I have been an early adopter or early developer of some technologies. I am confident doing my own research and even reaching out to the company when I need support to succeed at using it.
• Support from others: I don’t need anyone to learn it with me, but I enjoy it if they do. I like to learn many different ways to use technology so I can teach others and get them hooked on it….
• Life situations: I have a routine of things I like to do. I enjoy incorporating new ideas and technologies into my routines.
• Personality: I am open to experiences and perhaps extroverted. I look forward to overcoming challenges and difficulties. I enjoy building puzzles to help myself and other people become more familiar with the technology.
• Emotional connection: I see no reason to get emotionally connected to technology; it’s a lot of fun, but at the end of the day, it’s just a tool.
• Progress through acceptance phases: I speed ahead into late acceptance phases. When I start thinking of technology, I consider how to adapt to its possible strengths and drawbacks. Technical issues can still slow my use of the technology, but once I overcome them, I am quick to recover my progress and incorporate the technology into my routines.

4.2.3. Nora the Non-Starter

• Attitude: I wait until everyone else in my neighborhood gets the technology before I bother. I give technology exactly one chance to work. If it doesn’t, then I want to throw it out.
• Technology experience: Limited; I only use the basics like phone calls, emails, texting, Twitter (X), and Facebook messaging. I can be convinced to use technology if it has a functional use that I need, like with my Bluetooth hearing aid and off-brand robot vacuum cleaner.
• Support from others: I do not have people in my life to help me use new technology. This means that if I can’t figure out how to use it on my own right away, I probably won’t use it at all.
• Life situations: It takes other people to keep me going with using technologies in routines, especially if the routines require energy from me. If I don’t have this support, anything new will fade away.
• Personality: I am steadfast. Once I make up my mind about technology, the technology, and other people cannot change my mind. If other people think the technology is easy to use, but I can’t make it work, it is probably the fault of the technology, and not worth my effort trying to make it work.
• Emotional connection: I feel no emotional connection to most technologies. Some new technologies only take up space.
• Progress through acceptance phases: I have high expectations of technology, and if it does not work the way I think it should, I will only find disadvantages in it and discard it quickly, skipping to a Non Use phase without fully experiencing the technology. It is possible for me to form routines with the technology, but only if somebody else creates these routines and makes the technology automatically work for me.

4.3. RQ3: How to Identify Which Phases Are Occurring and Suggested Changes to Longitudinal Technology Acceptance Framework

Based on this research, we have several recommendations for improving longitudinal technology acceptance studies. We propose a new method for identifying acceptance phases, which acknowledges the fluidity of the phased framework of longitudinal technology acceptance. We have new recommendations for coding participant responses as fitting in each phase. Finally, we have recommendations for improved interview questions to help researchers keep from biasing participant responses toward certain acceptance phases.

First, we propose a new method for identifying which phase(s) participants are in. In prior work, researchers labeled phases based on the expected timeline and which phase was dominantly coded for the average participant at each given time [9,58,96,97]. In this research, we found that phases were not so clean-cut, with evidence of participants experiencing multiple phases at once and progressing through the phases in different orders. To better suit these data, we propose a new method of identifying which phase(s) each participant is in during a given session, based on observing “peaks” in phase codes. We refer readers to Section 3.1 for a detailed description of our novel method. This method allows for more nuance and precision in understanding individuals’ experiences.

We also have new recommendations for how researchers should code participant responses as fitting into a phase. For example, when participants noticed the disadvantages of the robots, this is currently coded as evidence for the Identification₆ phase. However, we found that often it instead signaled that people did not like the robot, would not use it, and might progress toward the disuse phase. Thus, we provide a more detailed coding guideline to help future researchers more accurately determine what acceptance phase(s) participants are in (Appendix E).

Finally, we recommend changes for interview questions for longitudinal studies on technology acceptance phases. Interview sessions should avoid questions that directly point to codes from the phases, as it will increase that phase’s appearance at various intervals rather than when it naturally appears for individuals. Future researchers should build upon the framework proposed by [9] and the framework included in Appendix E. These improvements in understanding phases will strengthen the future development of longitudinal technology acceptance theories.

5. Limitations and Future Directions

In this paper, we propose improvements to understanding acceptance phases, coding them, and asking interview questions. We recommend that future research test and validate these recommendations and continue making improvements to procedures for long-term technology acceptance studies.

Like with all research, there are a number of limitations. The results of this case study are limited to a small subject pool. With six participants in our original study, only four are included in this case study. All participants in our study pool are within the small age pool (60+). Most other studies of technology acceptance included more than twenty participants. All participants were recruited lived in New Mexico, and their primary language spoken was English. Future research should extend this work to broader groups in terms of ethnicity, location, socioeconomic status, digital literacy, health conditions, and age if research is conducted beyond the older adult population.

Additionally, the research presented in this paper is qualitative data based on transcripts of interviews throughout the period of time when participants hosted the robots in their homes. The robot is a commercially available telepresence robot (Double 3 [68]), therefore, direct logs about the mobile telepresence robot usage statistics and errors were not available to us. Future research should consider objective performance metrics that can be assessed in real-time to better support a user’s experience of, interaction with, and/or control of the robot. It is now feasible to utilize machine learning (ML) techniques to study technology acceptance outside of conventional social science research given the rise in popularity of ML applications in the last 5 years [98,99]. In the context of mobile telepresence robots where the goal of the technology is to enhance human-to-human interactions (not to optimize the interaction and/or interface between the user and the robot), there are a number of usage statistics that can be used to measure the quality of an interaction for mobile telepresence robots [100], including communication methods, quality of audiovisual interactions, failure rates, and time taken to resolve technical problems. These performance metrics have been used for short-term or duration deployments. These should be reconsidered for long-term deployments to better understand users’ behaviors and technology acceptance; for example, a long (duration) call could mean either the users are having a good conversation or alternatively that they are experiencing problems and trying to troubleshoot together. It will be important for future researchers to consider not only the type of data but also the sampling frequency necessary in order to achieve a sufficient amount of data for ML models. In this study, we conducted monthly interviews which resulted in too small of data for quantitative analysis.

Future research should look into various robots for technology acceptance to determine if the embodiment, character, or uses of the device influence acceptance. Researchers should dive into whether users prefer human-like interactions or robotic-like interactions. Researchers should better define what personas users tend to want which types of interactions and embodiment forms of robots.

6. Summary and Conclusions

In this seven-month longitudinal study, we introduced Double Robotics 3 mobile telepresence robots into the homes of older adults aged 60+. This case study included four participants from the original study [10] who had the most diverse responses to the robots. Participants completed monthly interviews about their uses, encounters, and acceptance of the robot over seven months. Our goal was to learn what hindered and what advanced technology acceptance and to advance a more rigorous definition of what acceptance phases participants experienced at what time, based on the data. We found that these phases are fluid as users can move between phases at various times and that phases can co-occur. Our novel contributions are: We propose a way to discuss participants’ experience of these fluid phases. Participants’ progression through acceptance phases was supported and hindered based on several factors, which relate to, but do not fully align with previous technology acceptance models. Our main factors were problem-solving with technology, life situations (especially business and routines), and overall personality or mindset. We recommend that researchers incorporate these into technology acceptance models, especially for long-term acceptance. We also created three data-based personas to help scholars keep potential users in mind when developing technology for older adults. Future research should continue to use and hone our fluid phased framework for technology acceptance and develop more user personas based on a broader population of older adults (e.g., from different locations and with different native languages).