An investigation into the use of smart home devices, user preferences, and impact during COVID-19

Abstract

With the goal of designing smart environments that can support users’ physical/mental well-being, we studied users’ experiences and different factors that can influence success of smart home devices through an online study conducted during and after the COVID-19 restrictions in June 2021 (109 participants) and March 2022 (81 participants). We investigated what motivates users to buy smart home devices, and if smart home devices may have the potential to improve different aspects of users’ well-being. As COVID-19 emphasized a situation where people spent a significant amount of time at home in Canada, we also asked if/how COVID-19 motivated purchase of smart-home devices and how these devices affected participants during the pandemic. Our results provide insights into different aspects that may motivate the purchase of smart home devices and users’ concerns. The results also suggest that there may be correlations between the use of specific types of devices and psychological well-being.

1. Introduction

Among many other benefits, smart home environments allow users to control the home environment using simple commands and can make users aware of possible problems inside their home environment. Voice assistants, virtual agents, and social robots can also provide users with some levels of companionship, and adding fun and enjoyment in controlling home environments. Therefore, they can be effective in improving quality of life and increasing independent living (Marikyan et al., 2019), and have a great potential to improve mental and physical health and well-being.1 Smart home devices have many advantages over other types of technologies, e.g., they are flexible to be added to different environments and the devices can be personalized based on each user's needs. Smart homes can include socially intelligent agents (e.g., social robots), which are shown to be effective in many health-related contexts (Chen et al., 2018; Marikyan et al., 2019; Martín et al., 2013; Saunders et al., 2015; Shishehgar et al., 2019), and can also increase accessibility and assistance capabilities of other smart home devices and systems. For example, aside from the sense of control (Demiris et al., 2008), security (Chalhoub et al., 2020; Pietrzak et al., 2014), and confidence (Pietrzak et al., 2014), older adults can benefit from even simple features such as light control to avoid falls. A smart home environment can also help younger adults to stay connected with an older relative, and to notify them about situations that may cause a threat to an older relative's health.

Despite these potential benefits, smart home devices are not used in many homes where they can be highly beneficial (e.g., for supporting older adults). In most cases, factors such as security and privacy (Mocrii et al., 2018), cost (Georgiev & Schlögl, 2018), user acceptance (Dragone et al., 2015), challenges with connectivity of devices, and even a lack of experience/knowledge about the potential benefits of these environments have been shown to affect their adoption. A step toward building a successful smart home environment is to understand how smart home devices are currently used and study the related preferences, concerns, and considerations for using smart home environments beyond the technical challenges such as connectivity. This study provides insights regarding these aspects.

With a long-term goal of building a smart home environment that can successfully support health and well-being, in this study we asked (1) how people of different ages with different personalities (Big 5 (Block, 1995)) and levels of emotional and psychological well-being use smart home devices/environments, (2) what would motivate users to purchase smart home devices and what tasks would benefit the most from automation, (3) what are the existing devices that have the potential to be used in an smart home environment, and specifically how they can support older adults or other users during a pandemic like COVID-19, and (4) what would be users’ potential concerns. As COVID-19 has increased social isolation among people of all ages, and has caused a situation where many people have spent a considerable amount of time at home and worked from home, we asked (5) if/how smart home devices affected people's lives during the pandemic, and if the pandemic motivated the purchase of a smart home device. We also further studied reasons for not having a smart home device, and, with cost of devices being one of the important considerations mentioned in previous work (Georgiev & Schlögl, 2018), we studied (6) what would be an acceptable price for a smart home device from participants’ point of view.

Other studies have investigated related issues. For example, Jensen et al. (2018) conducted a study to understand users’ needs and expectations from smart home devices with 23 participants in Australia and followed Nelson and Stolterman's concept of desiderata to understand people's experience and desires of smart home devices (Jensen et al., 2018). Also, Sing et al. (2018) conducted a study to understand people's opinions about and attitudes towards different types of smart home technologies (Singh et al., 2018). Further, Georgiev and Schlögl (2018) conducted an interview with smart home technology users to study the potential challenges in adoption of smart home devices (Georgiev & Schlögl, 2018). We build on these studies by investigating user needs, concerns, and motivations for buying smart home devices with a large group of participants, the majority of whom already own smart home devices, and are also involving a more diverse group of participants, especially in terms of household income, including those who do not have high household incomes. We hope that our findings can help designers and researchers to better understand the potential target users for specific technologies, as well as the important factors to consider and address in order to reduce users’ concerns and increase adoption and acceptance of smart home technologies.

2. Research questions

To better understand users’ needs, expectations, and concerns, and with the ultimate goal of designing smart home environments that can truly support people, we conducted an online study to address the following research questions.

• RQ1: What are the most popular smart home devices that users have at home (if any)?

• RQ2: What are the reasons for not having a smart home device?

• RQ3: Has the pandemic motivated purchase of smart home devices, and if/how using those devices affected users and their well-being during the COVID-19 Restrictions, as well as after those restrictions were lifted?

• RQ4: What motivates one to buy a smart home device, what functionalities do people find to be the most beneficial for smart home devices, and how much are they willing to pay for a smart home device that meets their needs?

• RQ5: What are people's opinions about the use of smart homes during the pandemic and for supporting older adults?

• RQ6: What are people's concerns about the use of smart home devices for supporting older adults and in homes with children?

Our study was exploratory in nature; however, some of these questions, specifically RQ4, and to some extent RQ2, have been studied in other countries (e.g., see (Singh et al., 2018; Wilson et al., 2017)), and we expect that results on motivations can, to some extent, be consistent with these previous findings (Singh et al., 2018; Wilson et al., 2017). Also, our research question about the pandemic was motivated by a recent study that showed a change in perception of socially intelligent agents (social robots) due to the pandemic in Canada (Ghafurian et al., 2021).

3. Related work

As mentioned before, Jenson et al. (2018) studied users’ desires regarding smart home systems. Results were discussed by creating three personas, the helper, the optimiser, and the hedonist. The helper represented the desired functional purpose (e.g., new desired functionalities and devices that consume energy). The optimiser was about the desired outcome (e.g., support assisted living and improve comfort and convenience), and hedonist was about the desired experience (e.g., new devices are bought because they are fun) (Jensen et al., 2018). Further, Wilson et al. (2017) studied how users perceive benefits and risks of smart home devices and discussed policy implications (Wilson et al., 2017). They found “managing energy use”, “controlling the domestic environment”, and “improving security” to be important functionalities and emphasized the importance of aspects such as cost, comfort, and convenience (Wilson et al., 2017). Differences in preferences/expectations of user perceptions and industry marketing were also emphasized (Wilson et al., 2017).

In an attempt to understand people's opinions about and attitudes towards different types of smart home technologies, Sing et al. (2018) conducted a study and found quality of life, safety, and comfort to be three factors for which the participants perceived that smart home devices would have the most potential benefits (Singh et al., 2018). Dependency on technology and privacy were two of the concerns indicated by the participants (Singh et al., 2018). Also, Georgiev and Schlögl (2018) conducted an interview with the users of smart home devices and found privacy and security issues to be the most important concerns, affecting adoptions of smart home devices, and emphasized the importance of interoperability of the devices (Georgiev & Schlögl, 2018). Also, access to information related to security and privacy was found to be another important factor. In a recent study, it was reported that the participants who sought information on privacy and security prior to the purchase of IoT devices found it to be difficult or impossible to find (Emami-Naeini et al., 2019, pp. 1–12). In this study, also, privacy and security concerns were ranked among the most important considerations for purchasing IoT devices, after the devices’ features and price (Emami-Naeini et al., 2019, pp. 1–12). Likewise, the importance of data transparency for increasing privacy was emphasized in a recent co-design study (Yao et al., 2019, pp. 1–12).

Other studies considered having multiple users, stakeholders, as well as different types of users in a home (Garg & Sengupta, 2020; Geeng & Roesner, 2019; Strengers et al., 2019). For example, Tabassum et al. (2020) investigated the role of stakeholders outside a home who have access to the smart home devices and suggested that individuals may provide remote access to trusted family and friends to gain more safety/security (Tabassum et al., 2020). It has also been shown that different users, e.g., primary and secondary users within a home may have different views about privacy (Lau et al., 2018). Also, it was found that IoT device owners may prefer for the device outputs to change in the presence of the visitors, and that the visitors would want to be aware of the type of data collected from them (Marky et al., 2020). Furthermore, a recent study (Ehrenberg & Keinonen, 2021) investigated how smart home devices affect the experience of home, and discussed ethical challenges as the results suggested that smart homes may create hierarchies within a household and constrain interactions (Ehrenberg & Keinonen, 2021).

In another study, Koshy et al. (2021) divided smart home users into pilots (users who introduce new features to their smart homes) and passengers (users who do not change features) and showed how smart home devices may be used differently by these two groups (e.g., how they acquire information for their devices, who they refer to for device-related assistance, etc.) (Koshy et al., 2021). Further, preferred locations in a home and activities that user prefers not to be recorded have been studied by Choe et al. (2011) (Choe et al., 2011). Also, personalization of smart devices in general has attracted researchers’ attention. Personalization has been achieved through different ways, such as by analyzing and learning from users’ data (Augusto et al., 2005), or by having devices that can directly learn from and/or being taught by users (Manca et al., 2020; Saunders et al., 2015).

Sense of control is another factor which was studied in the previous work. Luria et al. (2017) studied the sense of control for controlling smart home devices, and showed that sense of control may be higher when people use familiar interfaces to control smart home devices, and the lowest with voice control. They suggested the replacement of voice control systems with social robots (Luria et al., 2017). Social robots’ benefits over virtual agents has in fact been emphasized in the previous literature, e.g., suggesting that “social presence” is an aspect that can improve perception of social robots over virtual agents (and not necessarily their physical embodiment) (Li, 2015). Similarly, in a study involving co-design sessions on improving privacy, control was identified among one of six key design factors (Yao et al., 2019, pp. 1–12). Possible scenarios and potential benefits of using social robots within smart home environments have also been studied (Koay et al., 2020; Li et al., 2018; Saunders et al., 2015). For social robots, preference for the level of control was also shown to depend on the task that the agent performs (Chanseau et al., 2019).

Other studies have also discussed the potential benefits of using smart home devices in homes with children (Nicholls et al., 2020). Adoption of smart home devices has also been studied for older adults. Pal et al. (2018) studied adoption of older adults and found factors such as affordability and privacy to affect intention to use the smart home devices, while ease of use was found to be a factor that positively affected the perceived usefulness of smart home devices (Pal et al., 2018). Another study suggested that older adults have a generally positive attitude towards using smart home applications, especially for health monitoring and independent living, while being concerned about consequences such as social-isolation (Pal et al., 2017). Also, in a recent study investigating older adult users’ opinions about a voice assistant (i.e., Alexa), limited benefits/value was one of the reasons mentioned for limiting the use of (or even abandoning) the device (Trajkova & Martin-Hammond, 2020). This study suggested that the top three commonly discussed functionalities by older adults were (a) to listen to music/radio, (b) to set alarms/timers/reminders, and (c) to get information about the weather (Trajkova & Martin-Hammond, 2020).

Furthermore, for supporting older adults, smart home devices have been emphasized to be beneficial as facilitators of health and well-being interventions, while emphasizing the need of professionals in teaching new skills, installing devices, etc (Van Grootven & van Achterberg, 2019). Yet, an overview of 152 existing projects on the use of information and communications technology for older adults suggested that there is still no evidence of increased health and well-being outcomes in older adults in Europe (Van Grootven & van Achterberg, 2019).

Smart homes have also attracted attention during COVID-19. For example, methods have been proposed for remotely monitoring patients during quarantines (Taiwo & Ezugwu, 2020), as well as for prevention and control of COVID-19 (Dong & Yao, 2021; Gupta et al., 2021). Also, a recent survey found that those with a moderate perceived risk of getting COVID-19 were willing to pay more for smart devices with well-being features, as compared with those with high-risk or low-risk groups (Chen et al., 2020). This survey also suggested that one third of the participants were willing to pay for features that can reduce environmental impacts (Chen et al., 2020). Further, Brem et al. (2021) (Brem et al., 2021) provided an overview of how technologies in general may improve lives during COVID-19. Finally, Zanocco et al. (2020) showed that during COVID-19 restrictions (“Shelter-in-Place” order in California), those who reported a higher changes in occupancy and activity measures also reported a higher intention to adopt smart technologies (Zanocco et al., 2020). All of these studies could suggest that COVID-19 might have motivated use/purchase of smart home devices. Yet, it is not clearly how these devices might have influenced people's lives during the pandemic.

4. Methodology

A six-part questionnaire was designed to study the above-mentioned research questions:

• Section 1 - Loneliness: Loneliness was measured using the 8-item UCLA Loneliness Scale Questionnaire (ULS-8) (Hays & DiMatteo, 1987) 2

• Section 2 - Big 5 Personality: Big 5 personality, which has five different traits of Extraversion, Conscientiousness, Openness to experience, Agreeableness, and Emotional Stability (or Neuroticism) was measured through the TIPI Questionnaire (Gosling et al., 2003). We chose TIPI as a standard questionnaire with a reasonably short number of questions that is commonly used in studies in Human-Computer Interaction, and especially in the studies in Human-Robot Interaction.

• Section 3 - Use of and Opinion about Smart Home Devices: A questionnaire was designed to understand participants’ opinions about and experiences with smart home devices. Participants were first asked if they have any smart home devices. If their responses were yes, they were asked: (a) what type of devices they have, (b) if they think the devices positively affected them during COVID-19 and how/why, (d) if they bought the device during the COVID-19 situation, and if COVID-19 motivated their purchase.

If the participants did not have any smart home devices, they were asked: (a) why they were not using a smart home device, (b) if they were interested to have a smart home device, and (c) if they had ever seen or interacted with the smart home device, and which types of devices they had interacted with.

For the questions that asked about the type of smart home devices, the categories were decided based on the previous literature, i.e., according to (Jensen et al., 2018; Mekuria et al., 2019; Singh et al., 2018). The questions were multiple-choice options and the final categories were as below:

• ●Voice assistant agent or robot (e.g., Google Home, Alexa, Jibo, Vector, etc.)
• ●Vacuuming or Mopping Robots, or a robotic lawn mower (e.g., Rumba, Braava Jet, Shark IQ, etc.)
• ●Smart lights (e.g., Philips Hue, Wiz, Ring, etc.)
• ●Home security systems (e.g., security cameras, contact sensors, motion detectors, etc.)
• ●Cameras for monitoring people/pets (e.g., monitoring an older relative, a child, or even a pet)
• ●Smart smoke, carbon monoxide (co), or water leak detectors, etc. (e.g., Google Nest Alarm, etc.)
• ●Smart home appliances (e.g., smart fridge, such as a fridge with a tablet that either has voice assistants)
• ●Smart medical or health related sensors (e.g., respiratory rate monitor, sleep monitor, heart rate sensors, fall detection, etc.)
• ●Wearable devices (e.g., fitbit, smart watch, etc.)
• ●Other (please specify)

Next, all participants answered the following questions:

• ●
  “If you can get smart home devices to help you with different tasks around home, which tasks do you think would benefit the most from being automated?” Response categories were decided based on (Marikyan et al., 2019) and were:

  • –
    Comfort, for example, automation of daily routines, leisure, etc
  • –
    Health Monitoring, for example, wearable sensors, blood oxygen level, blood pressure, etc.
  • –
    Health therapy, for example, remote therapy, delivery of health-related services, etc.
  • –
    Companionship and Support, for example, companion virtual agents or robots, alarm systems for improving vision/hearing, robotic devices for rehabilitation
  • –
    Sensors that advise you about positive changes to make in your house, e.g., approaches to save energy, etc.
  • –
    Sensors that monitor changes in your house, e.g., water consumption, energy usage
  • –
    I do not want anything automated or to use any smart devices
  • –
    Other (please specify)
• ●“How much will you be willing to pay to have a smart home device for the functionalities that you selected above?” Options ranged from “A maximum of $50” to “More than $1000”. Participants could also choose “Other” and specify.
• ●“What motivates you/motivated you to buy a smart home device?” Response categories were decided based on (Singh et al., 2018), while we added three more categories that we believed would complement the categories suggested in (Singh et al., 2018). These categories were: (a) Save time, (b) Save energy, (c) Save money, (d) Provide comfort and make things effortless, (e) Provide peace of mind, (f) Improve quality of life, (g) Improve health, (h) Provide safety, (i) Provide companionship, (j) Increase property value, (k) My general interest in the latest technology, (l) To impress my friends and family, (m) Other (please specify), and (n) Nothing
• ●“Which of the smart home devices do you think would be the most beneficial during a pandemic, or for those who are living alone?” (options were similar to the list of devices provided above)
• ●“Which of the smart home devices do you think would be the most beneficial for older adults who live alone?” (options were similar to the list of devices provided above)
• ●“Do you have any concerns about using smart home devices in an older adult's home?” (answered in a text box)
• ●“Do you have any concerns about using smart home devices in a home that has children?” (answered in a text box)

• Section 4 - Attitude Towards Technology: As attitude towards technology can affect preferences and use of smart home devices, the attitude toward technology sub-scales from the Media and Technology Usage and Attitudes Questionnaire (MTUAS) (Rosen et al., 2013) were used. This standard questionnaire uses 16 questions to evaluate users’ attitude toward technology by providing statements such as “New technology makes people waste too much time” and “Technology will provide solutions to many of our problems”. These questions were answered using a 5-point scale, ranging from ”strongly disagree” to ”strongly agree”.3

• Section 5 - Demographics: Participants’ demographics information and other information related to general experiences, as well as their experience of COVID-19 was gathered. This questionnaire included: (a) age, (b) gender, (c) education level, (d) household income, (e) if participants had pets/type of pets, (f) number of people in their household, (g) number of children in their household, (h) how they felt about COVID-19 and social isolation (responses ranged from their life is completely and negatively affected to their life is completely and positively affected), (i) how stressed/anxious they were due to COVID-19, (j) how safe they feel at home (taken from (Singh et al., 2018)).

Also, in an attempt to encourage participants to end this part with a positive note (having asked about issues that might provoke negative thoughts, loneliness, experience of COVID-19, etc.), we asked them to state their favourite animal, favourite cartoon character, and favourite movie.

• Section 6 - Psychological Well-being: There are many tools for measurement of well-being (Linton et al., 2016). The Ryff's Psychological Well-Being (PWB) Questionnaire is a well-validated scale that has the advantage of including a reasonable number of items and of assessing several distinct factors generally considered central to well-being. Therefore, the Ryff's Psychological Well-Being (PWB) Questionnaire was used in this study to measure psychological well-being. This questionnaire has 42 items rated on a 6-point scale ranging from “strongly disagree” to “strongly agree”. Ryff's PWB measures five aspects of well-being (Ryff, 1989):

• ●Autonomy: Being independent and self-determined.
• ●Environmental mastery: Feeling competence in managing the environment and controlling external activities (Ryff, 1989).
• ●Personal growth: Seeing self as growing, being open to new experiences, and feeling realizing one's potentials (Ryff, 1989).
• ●Positive relations: Having satisfying relationships with the others and being concerned about their welfare (Ryff, 1989).
• ●Purpose in life: Having goals in life and holding beliefs that give life purpose (Ryff, 1989).
• ●Self-acceptance: Having a positive attitude about self and acknowledging both good and bad qualities (Ryff, 1989).

Attention and sanity checks were added on different pages of the questionnaire (e.g., “How much do you think that drinking water is liquid”). Also, participants could choose not to answer any question on the different pages of the questionnaire, as required by the institution's Research Ethics Board.

4.1. Procedure

After reading and accepting the consent form, participants were directed to the questionnaire and completed all 6 steps. They were given a completion code and instructions on how to submit the Amazon Mechanical Turk HIT4 afterwards.

4.2. Data analysis

As the categories for smart home devices were selected to be clear for the participants, we used them in the questionnaire, but grouped them for studying effects on well-being based on (a) giving control (including vacuuming/moping/lawn mower robots, smart appliances, and smart lights), (b) providing security/safety (including home security systems, cameras, and smart some/carbon/water leak detectors), (c) having social benefits (voice assistants), and (d) supporting health (wearable devices and smart medical/health related sensors).

Linear regression models were used to study significance while controlling for the confounding factors. Binomial tests were used for comparing significant differences between selecting two choices. Thematic analysis was used to analyze the qualitative data (i.e., opinions and comments about a specific choice). Themes were coded by hand. We followed the guidelines provided in (Gregar, 1994) for analyzing and reporting qualitative data.

4.3. Participants

Participants were recruited at two different points in time. Once in June 2021 (called Study 1 thereafter), when the safety measures for COVID-19 were in place in Canada and followed, and once in mid March to April 2022 (called Study 2 thereafter), when many of those measures were lifted. We ensured that the participants joining in 2022 have not joined the study before in 2021. A total of 198 participants were recruited. Data from 8 participants were removed from the study as they failed the attention checks. This left a total of 190 participants, 109 recruited in 2021 (will be called Study 1 thereafter), and 81 in 2022 (will be called Study 2 thereafter).5 74 reported to be female, 114 male, and 2 unknown (age range [18,67]; average: 35.6 yrs).

Participants were recruited on Amazon Mechanical Turk. Participation was limited to those who had an approval rate over 94% and had completed at least 50 HITs. This criteria was selected according to the previous studies and to ensure that the participants were familiar with the interface and to increase the chance that the responses will be reliable. Participation was limited to [COUNTRY]. The study was estimated to take up to 30 min and the participants received 4 USD for completing all steps of the study, and a pro-rated amount if they wished to stop earlier.

Fig. 2 shows the distribution of participants’ age. Fig. 1 (a) shows participants’ reported income, suggesting a diverse group of participants. We are also reporting on the income of participants who have a smart home device in Fig. 1 (b), and for those without a smart home device in Fig. 1 (c), which can be informative for understanding the two participant groups and the responses given by each group (e.g., reasons for not having a smart device).

Fig. 2.

Distribution of age of (a) All participants, (b) participants in Study 1, and (c) participants in Study 2.

Fig. 1.

(a) All participants’ income, (b) income of participants who had a smart home device, and (c) income of participants who did not have a smart home device.

Full Ethics clearance was received from the University of Waterloo's Human Research Ethics Board prior to running the study. The study was conducted in June 2021. Participants were informed about the nature of the questions under “foreseeable risks” in the consent form.

5. Results

142 participants reported that they had at least one smart home device, and 48 participants did not have any smart home device. 61 participants (43% of those with a smart home device) reported that they had bought a smart home device during the pandemic.

In this section we will first start by discussing the results related to the overall use and perception of smart home devices (RQ1, RQ2, RQ4, and RQ6). Afterwards, we will discuss the results related to smart home devices and COVID-19 (RQ3 and RQ5).

5.1. Overall use and perception of smart home devices

5.1.1. Participants with smart home devices - RQ1

Types of devices currently owned by the participants can be found in Fig. 3 . Voice assistant agents/robots were the most commonly owned devices, which the participants owned significantly more than all other types of smart devices (p < .001). It was followed by wearable devices and smart lights, both of which were owned more than the other devices (p < .05), except for voice assistant agents/robots. Smart medical or health sensors were owned significantly less than all other devices (p < .01, owned by only four participants).

Fig. 3.

Participants’ Existing Devices. Significant differences are shown on the plot with lines and stars. Note that in this figure and the other similar figures in this paper, a line indicates significant difference between any category on the left end of that line and on the right end of that line. As the bars are ordered by number, this would also mean that any category on the right of the line are also significantly lower than any category on the left side of the line. For example, this figure shows that smart lights were selected significantly more than home security systems, as well as any other category on the right side of home security systems. Also, as home security systems were selected significantly less than smart lights, they were also selected significantly less than any category on the left side of that, i.e., wearable devices and assistant agents/robots. * = p < .05, ** = p < .01, and *** = p < .001.

5.1.2. Participants without smart home devices - RQ2

Fig. 4 shows the selected reasons for not having a smart home device. Most participants who did not have smart home devices were concerned about their cost, followed by having privacy and security concerns. Not being familiar with smart home devices was selected significantly less than all other reasons (p < .05 to p < .0001), and cost was the option selected the most, and significantly more than all options other than “privacy and security”. Note that the size of data in this category was limited, and therefore may not be able to capture some significant differences.

Fig. 4.

Reasons for not having a smart home device as selected by the participants. * = p < .05.

Participants who did not have a smart home device were asked whether they would be interested to have a smart home device and to comment on it. The comments were first tagged as “positive”, “negative”, and “neutral”, showing participants’ general interest in having a smart home devices. For those that were tagged as positive or negative, thematic analysis was done on the reason. Below are examples of each category:

• ●Positive:“From what I've heard from friends and family who do use smart home devices, they say that it has helped to simplify their lives a great deal.”6
• ●Negative:“I'm not sure how much benefit I will get from these devices so I am hesitant to spend the money to purchase one to try out.”
• ●: Neutral: “I'm indifferent. I don't think I would purchase one myself, but if I were ever gifted one then I'd use it.”

Three participants’ comments were tagged as both positive and negative as they contained both a positive and negative component (e.g., ”I want them for the convenience but not sure how reliable they are or how costly they are (not just for purchase, but also for maintenance). ”). In addition to these three, 18 comments were tagged as positive (11 in Study 1 and 5 in Study 2), which had clear indications that the participants would be interested in buying a smart home device or where the participant only emphasized the benefits. Thematic analysis showed that devices for (a) comfort, to make tasks easier or help save time (e.g., a vacuum robot helping with cleaning; comfort was mentioned by 14 participants, eight in Study 1 and six participant in Study 2), followed by (b) safety/security (e.g., security cameras; safety/security was mentioned by 3 participants all in Study 1) were the most commonly mentioned devices that people without a smart home device wished to have/buy.

• ●Comfort:“In some cases, for example a vacuum robot, the smart home devices are very useful. But I am not sure I can afford one at this time. So I'm planning to buy one ASAP.”
• ●safety/security:“I would love to have a video camera doorbell, like Ring, but I fear that it is not as secure as we may think it is. I also have privacy concerns with voice-assistants like Alexa, Siri, etc.”

Further, in addition to the three comments that were tagged as both positive and negative, 24 comments (12 in Study 1 and 12 in Study 2) were tagged as negative. Thematic analysis showed (a) cost (mentioned by five participants in Study 1 and two in Study 2) and (b) privacy/security concerns (mentioned by five participants in Study 1 and four participants in Study 2) were the most commonly mentioned reasons for not being willing to purchase a smart home device.

Furthermore, Fig. 5 shows the devices that the participants without a smart home device had interacted with. This shows that the majority of the participants without a smart home device were familiar in a way with these devices by interacting with them elsewhere. Note that although we show significant differences in Fig. 5, due to the limited number of participants in each of these categories, it is not possible to draw any firm conclusions from these differences.

Fig. 5.

Devices that participants without a smart home device had seen and interacted with.

5.2. Tasks, motivation, and reasonable price to purchase smart home devices - RQ4

Fig. 6 (a) shows the results for the reported motivations for purchasing smart home devices by the participants, and Fig. 6 (b) shows the tasks the participants thought would benefit the most from being automated. Significant differences among different choices are shown on both figures.

Fig. 6.

(a) Motivation for purchasing smart home devices as reported by the participants. (b) Tasks (task goals) for which the participants were most willing to purchase a smart home device. * = p < .05 and *** = p < .001.

Despite that many participants reported to have a voice assistant agents/robots, “providing companionship” was one of the reasons that was selected significantly less than the others, along with “increasing home property”, and “impressing friends and family”. Saving time was selected more than the other options, and significantly more than any reason other than improving “quality of life” and “providing comfort and making things effortless”.

As for the tasks, providing comfort was selected significantly more than all the other choices. Therapy and providing companionship were selected significantly less than the others. Finally, a few participants indicated that they were not willing to buy any devices.

Finally, Fig. 7 shows the distribution of costs that the participants were willing to pay for a smart home device, which meets their needs.

Fig. 7.

The cost for a smart home device that the participants without smart home devices reported to be willing to pay.

5.3. Concerns - RQ6

As mentioned before, participants were also asked about their concerns related to using smart home environments for older adults, and in homes with children. Thematic analysis was conducted on both comments to identify the concerns. Fig. 9 summarizes these findings.

Fig. 9.

(a) Concerns about the use of smart home devices for supporting older adults and (b) Concerns about the use of devices in houses with children. * = p < .05, ** = p < .01, *** = p < .001.

5.3.1. Concerns related to older adults

Thematic analysis on the concerns of participants about using smart home devices for supporting older adults resulted in finding four themes as below, ordered based on the number of times repeated in the comments:

• ●Ease of use and maintenance: this was the most commonly mentioned concern for older adults. This theme appeared in responses of 76 participants. Examples: “My only concern is that an older adult might have difficulty getting the devices to do what they are asking.” and “The chances that they might not know how to operate and troubleshoot the smart device.”
• ●No concerns: 72 participants indicated that they did not have any concern for using smart home devices for supporting older adults. Examples: “No, I feel that the learning curve for most smart device can be achieved by seniors, like my parents for example. With patience and some adjustments I feel that they'll get the hang of such devices.” and ”No, I believe the benefits are greater than any possible concerns.”
• ●Security and privacy: as detected in 28 participants’ comments, one of the concerns were about the security and privacy of using smart home devices. Example: “Yes, poor security and privacy issues related to commonly available smart home devices are concerns, especially in older adult's homes and in vulnerable populations that don't necessarily understand or are unable to address/fix these issues.”
• ●Safety - tripping hazard: this concern was mentioned by five participants, especially for devices that move around such as moping and vacuuming robots. Example: “If it breaks; especially a moving device-like a mopping and vacuuming one, because an older adult might trip over it.”

5.3.2. Concerns related to homes with children

Thematic analysis of participants’ comment about their concerns related to the use of smart home devices in homes with children found five themes as below (ordered according to how often they appeared in the comments):

• ●No concerns: 77 participants had no concern for using smart home devices in homes with children. Examples: “No, I do not since many smart home devices are designed in a way that will make it safe for all ages to use or not be able to use.” and ”Not really. Children can adapt really fast with new technology. They proof it already with smart phone and tablets that they can use at a really young age.”
• ●Being used in a way that is not intended: one of the commonly mentioned concerns was about children using technologies in a way that is not appropriate or a way that they should not use it. This theme appeared in comments of 55 participants. The concerns were about different consequences of using the devices in a way that is not appropriate, for example (a) the possibility of breaking the technology (22 participants), (b) using devices to put online orders (11 participants), (c) using devices to control the home environment or change settings, e.g., changing thermostat (7 participants), and (d) accessing inappropriate content (7 participants). Examples: “Children could learn to control different aspects of the house by accident. It could be too much power for a child to have.”, “I'd be worried kids could get access to something like my Amazon account and place a ton of orders for candy and toys.”, or “I'm concerned the children would access information that is not suitable to their age range.”
• ●Privacy and security: 34 participants had privacy and security concerns about using smart home devices in homes with children. Example: “Possibly using a camera and having hackers access it and have access to a video feed of my child.”
• ●Safety: 18 participants’ comments showed safety concerns. Examples: “The only concern I have is things that automate chores and such may be dangerous for smaller children.” and ”If the device has got hazards like electricity hazard, I will be really afraid to use it when there are children around. Unless it is designed in a way that is safe or should be placed out of reach.”
• ●Limiting/affecting children's interactions and increasing their reliance on technology: this theme was detected in responses of 7 participants. Examples: “I would worry that these devices would teach children to be completely reliant on machines - for example, not knowing how to look up a fact in a reference book because Alexa is always there.” and “I think for children, I might worry that they spend all day long on electronics, and miss out on experiencing the real world. If robots are doing the household chores, they don't learn what hard work feels like, and they also miss out on valuable physical activity. If they're indoors staring at a screen all day, it's bad for their health.”

5.4. Smart home devices and COVID-19

5.4.1. Benefits of smart home devices during the pandemic - RQ3

Participants who had a smart home device were asked if/how it helped them during the pandemic, and if the pandemic motivated purchase of smart home devices. A total of 61 participants (28 in Study 1, and 32 in Study 2) indicated that they bought a smart home device during the pandemic.

Benefits during the pandemic: Thematic analysis of participants’ comments resulted in finding six main as shown in Fig. 10. These themes are discussed below, ordered based on the number of times they appeared in the comments:

• ●Smart devices did not make any difference during the pandemic. This theme was found in responses of 44 (25 in Study 1 and 19 in Study 2) participants. Example: “Nothing really changed. I was home more during the pandemic so it was good to have these kind of services around. But I never truly depended on it differently because of the pandemic.” or “I don't think it enhanced my life specifically during the pandemic, but it is convenient at all times.”.
• ●Smart home devices increased comfort and helped with making things easier/more efficient. This theme was found in 43 responses (24 in Study 1, and 19 in Study 2). Example: “I Google Home optimizes and makes certain tasks a lot easier at home. Especially because I spent most of my time at home during the pandemic, and my office space was here. Overall the benefits of a Google Home made my pandemic life more comfortable and easier. That being said, I would still prefer to work in person.” (sic)
• ●Smart devices provided social companionship and increased social engagement. This theme was found in responses of 25 participants (11 in Study 1 and 14 in Study 2). Examples are: “Because it gave me a sense of actually having someone to talk to. Especially with the assistant. The assistant definitely helped me keep my sanity during the 3 month period where I was completely alone” and “During the lockdowns, I was dealing with various mental health issues and some assistants were recommended to me by a therapist so I could communicate with a bot without any judgement. This greatly helped me in improving my mental state.”
• ●Smart devices added fun (e.g., playing music). 23 participants’ responses (14 in Study 1 and 9 in Study 2) reflected this theme. Example: “We really got them just after lockdown started last year and learned about them. They are fun and enjoyable to play with. Very convenient for music and grocery lists as well as lights and temperature control. They are helping us save money even though we are home all the time now. Also for shopping during covid, knowing exactly what you need makes the trip more efficient and calls for fewer trips to the stores.” 7
• ●Smart devices helped to stay fit and encouraged doing exercises. This theme was identified in 13 participants’ comments (7 in Study 1 and 6 in Study 2). Example: “I am very active, but having a fitness watch helped me push myself and continue with my workouts when gyms/classes were closed. It helped a lot to see my heart rate during workouts and to get reminders to move.”
• ●Smart devices helped increase safety and security. This theme was observed in responses of 6 participants (5 in Study 1 and 1 in Study 2). Example: “Having a home security system was helpful with speaking with individuals who were at my front door and not having to come in contact with them.”

Fig. 10.

How smart home devices helped participants during the pandemic as reported by them. * = p < .05.

Impact of COVID-19 on Purchase of Smart Home Devices: While the pandemic was not a significant predictor of purchase of smart home devices, 30 out of 61 participants who purchased a smart home device during the pandemic reported that the pandemic motivated this purchase. In a thematic analysis of the comments, “work from home/spending more time at home” (10 participants), “staying fit, exercising, or health monitoring” (7 participants), and“being bored or alone” (6 participants) were the three most frequently mentioned reasons for why the pandemic motivated the purchase of a smart home device. Examples of comments for each theme is below:

• ●”I purchased an apple watch to help monitor possible symptoms of covid and give me a better sense of what is going on with myself and body”
• ●I was irritated by the pet hair on my floor because I was home during the day and saw it more so I bought a Roomba.
• ●”I was so bored - and sedentary - during the early stages of the pandemic (with all sporting activities shut down in my area), that I had to try something to stay fit. So yes, I have to credit COVID-19 with spurring me on to better health. :) ”

Those whose purchase was not motivated by the pandemic mentioned that they had the plan to buy a device already before the pandemic, or another reason (e.g., a recent accident or moving to a new house) motivated their decision.

5.4.2. Correlations between well-being and device type - RQ3

For addressing this research question, we analyze the data from Study 1 and Study 2 separately, as they were collected at different times: Well-being in study 1 was measured when COVID-19 restrictions were in place and people were more isolated. Data collection in Study 2 took place after everyone had a chance to be fully vaccinated and the COVID-19 restrictions were widely lifted. Here, we will first discuss the results of Study 1 and Study 2 separately and then compare them.

We studied correlations between having different types of smart home devices, categorized according to their function as explained before (i.e., giving control, providing security/safety, having social benefits, and supporting health), on Ryff's PWB's five measures of well-being (i.e., Environmental mastery, personal growth, positive relations, purpose in life, and self-acceptance). It is important to emphasize that this investigation was exploratory in nature, and while we controlled for a range of different potential confounds (e.g., participants’ age, personality type, gender, etc.), a correlation may still not necessarily mean that a specific smart home device was the reason/only reason for a higher level of well-being. Below are the results for Study 1.

Life purpose: Having a device with social benefits (e.g., voice assistant agents/robots) was a significant predictor of life purpose according to a linear regression model that also accounted for the other confounding factors such as Big 5 personality traits. Results are shown in Table 1 . Those who had a smart home device with social benefits had a significantly higher measured well-being on the aspect of “life purpose” (se = 0.821, t = 2.032, p < .05). Big 5 Emotionality (se = 0.333, t = 3.790, p < .001), extraversion (se = 0.294, t = 4.256, p < .0001), and conscientiousness (se = 0.417, t = 4.195, p < .0001) traits all positively affected this measure, leading to a higher “life purpose”.

Table 1.

Linear Regression model predicting the “life purpose” measure according to Big 5 personalities and having different categories of smart devices. SocialTRUE shows having a device with the aforementioned social component. Other predictors were added, but removed as they were not significant and did not improve the model fit according to the Akaike information criterion (AIC).

Covariate	Estimate	SE	t	Pr $(>|t|)$
Intercept	8.215	2.146	3.827	¡.0001
SocialTRUE	1.668	0.821	2.032	¡.05
Emotionality	1.263	0.333	3.790	¡.001
Extraversion	1.250	0.294	4.256	¡.0001
Conscientiousness	1.751	0.417	4.195	¡.0001

Environmental mastery: Having a device that improve security/safety (e.g., security cameras) was a significant predictor of environmental mastery according to a linear regression model that also accounted for the other confounding factors such as Big 5 personality traits. Results are shown in Table 2 . Those who had a device capable of increasing safety/security had significantly higher measured well-being on the aspect of “life purpose” (se = 0.834, t = 1.998, p < .05). Big 5 Emotionality (se = 0.305, t = 3.767, p < .001), extraversion (0.273, 3.033, p < .01), and conscientiousness (se = 0.380, t = 2.238, p < .05) traits all positively affected this measure.

Table 2.

Linear Regression model predicting the “environmental mastery” measure according to Big 5 personalities and having different categories of smart devices. SecurityTRUE shows having a device that provides safety/security, as discussed above.

Covariate	Estimate	SE	t	Pr $(>|t|)$
(Intercept)	14.132	1.968	7.182	$<.0001$
SecurityTRUE	1.666	0.834	1.998	$<.05$
Emotionality	1.149	0.305	3.767	$<.001$
GenderMale	−0.696	0.760	−0.915	0.362
Extraversion	0.827	0.273	3.033	$<.01$
Conscientiousness	0.851	0.380	2.238	$<.05$

Self acceptance: Having a device that increases control over the environment (e.g., smart lights) was a significant predictor of self acceptance according to a linear regression model that also accounted for the other confounding factors such as Big 5 personality traits. Table 3 summerizes the results. Those who had a device that increased their control over their environment had a significantly higher measured well-being on the aspect of “self acceptance” (se = 0.834, t = 1.998, p < .05). It was interesting that we did not see this effect on environmental mastery. Both Big 5 Emotionality (se = 0.377, t = 7.149, p < .0001) and extraversion (se = 0.392, t = 4.486, p < .0001) traits positively affected this measure.

Table 3.

Linear Regression model predicting the “self acceptance” measure according to Big 5 personalities and having different categories of smart devices. ControlTRUE shows having devices that provide control, as explained above. Other predictors were added, but removed to improve the model fit according to AIC.

Covariate	Estimate	SE	t	Pr $(>|t|)$
Intercept	6.713	1.961	3.423	$<.001$
ControlTRUE	2.278	1.109	2.055	$<.05$
Emotionality	2.698	0.377	7.149	$<.0001$
Extraversion	1.758	0.392	4.486	$<.0001$

Study 2 Results and Comparison with Study 1: In general, participants in Study 2 had a higher average score for all aspects of well-being (i.e., autonomy, environmental mastery, personal growth, positive relations, and self acceptance). This difference was significant for autonomy (t = −2.242, p < .05).

We modelled Study 2 data similar to the modelling discussed for Study 1. We no longer observed any of the effects found in the Study 1 data. For example, we no longer observe the effect of having a device with social benefits on the measured life purpose aspect of well-being (se = 1.123, t = 0.148, p = .883). And there is no effect of having a device improving safety/security on environmental mastery (se = 0.881, t = 0.054, p = .957). Similarly, we do not observe an effect of having a device that increases control on the environment on the self acceptance aspect of well-being (se = 1.356, t = 0.141, p = .888). While we cannot draw strong conclusions, this difference along with the similarity in the other aspects of results (suggesting that the population was similar between the studies) could show that the effect of devices on people's well-being could change and depend on e.g., situations such as COVID-19 restrictions that could make people more isolated. Table 4 shows an example of a model including study as a factor, showing these differences, i.e., that having a device with social benefits significantly affected “life purpose” in Study 1, but not in study 2.

Table 4.

Linear Regression model predicting the “life purpose” similar to the model presented in Table 1, while including study as a factor.

Covariate	Estimate	SE	t	Pr $(>|t|)$
Intercept	8.758	1.755	4.990	$<.0001$
SocialTRUE	1.722	0.866	1.988	$<.05$
StudyTwo	0.902	1.002	0.901	0.369
Emotionality	1.557	0.275	5.659	$<.0001$
Extraversion	0.942	0.233	4.049	$<.001$
Conscientiousness	1.581	0.323	4.895	$<.001$
SocialTRUE:StudyTwo	−1.692	1.328	−1.274	0.204

5.4.3. Devices beneficial during the pandemic and for supporting older adults who live alone - RQ5

Fig. 8 (a) shows the devices that were rated to be most beneficial for supporting older adults. Smart medical or health sensors were identified as the most beneficial devices, followed by those that are intended to increase safety (e.g., home security systems) and assist with daily tasks (e.g., vacuuming robots). Wearable devices, smart lights, and smart home appliances were rated as least beneficial for older adults.

Fig. 8.

(a) Devices rated to be most beneficial for supporting older adults and (b) Devices rated to be most beneficial for supporting people during the pandemic. * = p < .05, ** = p < .01.

The results for devices found to be beneficial during a pandemic were different, as expected. For this category, voice assistant agents/robots were selected more than all other options, and significantly more than every option other than smart medical and health sensors.

6. Discussion

In this paper we asked what are the most popular smart home devices that users in Canada own, and how these devices affected people during COVID-19. To better understand what motivates/discourages people to buy smart home devices, we also asked about the reasons for not having smart home device, what motivations to purchase smart home devices, the desirable functionalities, and accepted costs for the devices. Further, we asked which devices are believed to be most beneficial for people during the pandemic and older adults who live alone. To further understand important considerations, we also studied concerns related to using smart home devices in older adults’ homes and in homes with children. The study was conducted in two different time stamps, once during the COVID-19 lockdowns, and once later after COVID-19 vaccinations, when the social isolation rules were not as strict. Being an exploratory study and given the multi-faceted nature of our results, we decided not to discuss them sequentially and rather provide a more holistic discussion of the findings. To the best of our knowledge, this was the first study to investigate correlations between well-being and owning a device during COVID-19, as well as to explore current use, opinions, and preferences for using smart home technologies in Canada.

Our results suggested that voice assistant agents/robots were the most commonly owned devices, followed by smart lights and wearable devices. It is important to emphasize that many factors could have affected these results beyond their perceived usefulness and need. For example, the cost of the devices might have affected purchase of some devices such as smart home appliances. Yet, smart medical and health sensors — the costs of which are comparable to the other categories of the devices that were owned by many participants — were the category that was owned the least. Although we had 51 middle-aged and older adult participants, 35 of whom already owned a smart home device (only two were older adults, i.e. $>=$65 yrs; both owned smart home devices), these results may be in part affected by the age of participants. An alternative explanation is that the benefits of these devices might not have been clear to the users, and if that is the case, as these devices have potential to improve health in many domains, it would be beneficial to investigate approaches that inform users about the benefits of these devices in the future.

An interesting result was that although voice assistants were the most commonly owned devices, participants rated “providing companionship” as one of the least important motivations for purchasing the smart home devices. As also suggested in many comments, voice assistant agents/robots were used for activities such as listening to music or getting information faster. It would be interesting to study how voice assistants can provide companionship and if this aspect can be improved in the future, e.g., by increasing their social capabilities. Also, it is reasonable to assume that many participants had a voice assistant such as Google Home or Alexa, and fewer might have had a social robot. Therefore, it would be interesting to study if the use of an embodied agent such as a social robot would improve perception of companionship and strengthen this aspect as a motivation for purchasing smart home devices. On the other hand, even with social robots, it was previously shown that people would prefer the roles of assistant, machine, and servant more than a mate or friend for a robot (Dautenhahn et al., 2005); however, this preference might have changed over time.

For the pandemic, these two categories (i.e., voice assistant agents/robots and smart medical/health sensors) were found to be the most beneficial. Voice assistant agents/robots was the category that was selected the most by the participants. One explanation could be that, similar to what was observed for social robots (Ghafurian et al., 2021), the pandemic might have emphasized situations where people are socially isolated and emphasized the benefits of socially intelligent agents that provide a type of social presence. Health and smart sensors were the second most beneficial devices considered for the pandemic. It might be due to assumptions made regarding the expected or hypothetical functionalities of these devices, for example, thinking that they can detect symptoms of COVID-19.

Also, for supporting older adults, smart medical and health sensors were selected to be the most beneficial by the participants, while they were not popular among middle-aged or older adult participants. Also, devices that provided security, helped with domestic activities, and provided companionship (i.e., home security systems, vacuuming/moping robots, smoke/co/water leak detectors, and voice assistant agents/robots) were perceived to be equally beneficial for supporting older adults. These ratings are based on opinions of all participants (regardless of age), who were younger or middle-aged adults (except for two participants who were over the age of 65 years and could be categorized as older adults). Future studies would benefit from obtaining these ratings specifically from older adults to understand if and how these ratings may change when the participants are the primary target users; however, ratings from younger and middle-aged adults are still informative as past research has shown that family members can also influence older adults’ attitudes toward technologies (De Regge et al., 2020).

Understanding reasons for not having a smart home device can inform us about the aspects that can be improved to motivate users to buy a device, in order for them to take advantage of their benefits. Based on our findings, cost was the most commonly mentioned reason, followed by privacy/security concerns and uncertainty about the benefits/effectiveness of these devices. It is important to note that 11 out of 28 participants who did not own a smart home device showed a positive attitude towards them, and emphasized that they would be interested to own devices, especially for increasing their comfort and security. This suggests that by making smart home devices more affordable and more secure, and by informing people about the benefits of smart home devices, one might reach out to a larger group of people who could benefit from the use of smart home devices. While cost was one of major concerns and despite the diverse range of incomes, majority of the participants were willing to pay up to $100 and more than half (62%) were willing to pay up to $200 for purchasing smart home devices that met their needs.

In terms of functionalities for smart home devices, we found increasing comfort, monitoring changes in a home, advising on positive changes, and health monitoring to be considered as most beneficial functionalities for which the participants were willing to use (and pay for) a smart home device. It was interesting that health monitoring was among these choices although such devices were only owned by one participant. Further, some of our results on motivation for purchasing a smart home device were in line with findings of Sing et al. (2018) (Singh et al., 2018) and Wilson et al. (2017) (Wilson et al., 2017). For example, providing comfort and improving quality of life were two of the top three motivations mentioned for purchasing a smart home. However, unlike the findings of (Singh et al., 2018) and (Wilson et al., 2017), we found that saving time was rated as the most important motivation, which was selected significantly more than “providing safety” (a motivation that was found to be highly important and among the top three in (Singh et al., 2018)) and “save energy” (a motivation that was found to be most important in (Wilson et al., 2017)). While these differences might be due to differences between the countries in which the studies were conducted, and the demographics, it would be interesting for future work to study if/how COVID-19 affected these motivations, especially because making tasks easier was one of the factors emphasized by our participants about using smart home devices during the pandemic (e.g., when working from home).

With COVID-19 creating a situation where most people stayed at home, worked from home, and had limited social interactions, we studied the impact of having a smart home device on participants during COVID-19, and whether the pandemic motivated purchase of a smart home device. While 23 out of 81 participants reported that smart home devices did not make any difference during the pandemic, the majority of the participants emphasized that having a smart home device was helpful during the pandemic, and (a) increased comfort and made things easier/more efficient, (b) provided social companionship,(c) added fun, (d) helped them stay fit and motivated them to exercise, and (d) helped increase their safety/security. Although future work is needed to verify this, it is reasonable to assume that these benefits can go beyond the pandemic and be experienced in similar situations where people have to spend a significant amount of time at home and when human contact is limited or not possible (e.g., for older adults who may be socially isolated or those with disabilities).

To further study these benefits, we studied correlations between having specific categories of smart home devices and different measures of well-being according to the PWB standard questionnaire. We found having specific devices to be significant predictors of different measures of well-being, studied through linear regression models where potential confounding factors such as users’ demographics and personality types were controlled for. Having a device with social capabilities (voice assistant agents/robots) was a significant predictor of “life purpose”. Having a device that provided security (e.g., security cameras) was a significant predictor of “environmental mastery”, and having a device that provided control (e.g., smart lights) was a significant predictor of “self acceptance”.

An important point to emphasize is that the findings about the impact of smart home devices with social benefits on well-being may not actually be due to their social capabilities (although the social aspects were perceived to be important during the pandemic), rather this impact might be due to these devices’ abilities to help with organizing tasks, saving time, and increasing efficiency; especially because based on participants’ ratings, providing companionship was not found to be an important reason for purchasing smart home devices. In other words, the voice assistant agents/robots might have led to an increase in the “life purpose” measure of well-being because they helped manage people's lives and made tasks more efficient. Another explanation could be that those with a higher “life purpose”, i.e., those with more well-defined goals in their life, may be more interested in purchasing smart home devices that could increase their efficiency. Also, while all results were consistent across the two studies (i.e., two times that the study was conducted), we observed some differences in correlations between having these devices and well-being aspects between the two studies (see the Results section for more details). Overall, these results were a first step towards understanding how having different smart home devices may be correlated with different aspects of well-being and future work is needed to verify these findings through pre/post studies.

Finally, as expected, participants had multiple concerns related to use of smart home devices for supporting older adults and in homes with children. These findings can be informative for designers, emphasizing factors that should be taken into account to ensure that these devices are inclusive and can be adopted by a large range of users. For example, the findings emphasized that ease of use/maintenance is an important factor for smart home devices (similar to the other technologies designed for older adults); therefore, smart home devices can highly benefit from co-design with older adults. Also, as children may be secondary users of these technologies, it is important to consider concerns related to use of these technologies in homes with children when designing smart home devices. For example, inappropriate use of smart home devices by children was one of the major concerns of participants. This emphasizes the importance of adding, e.g, proper child locks and controls that would limit children's access to these technologies.

7. Limitations and future work

Our study had multiple limitations. Most importantly, while we controlled for different factors such as participants’ demographics and personalities, we cannot be sure that the observed effects on well-being were due to (or even solely due to) owning specific smart home devices. Future work should validate these results by studying the impact of smart home devices on well-being through before/after studies. Also, while the choice categories in all questions were based on extensive reviews of the literature and were decided according to previous studies and literature reviews, there is a possibility that specific categories were missed. To reduce the possibility that this would affect our findings, we provided participants with an “other” option to write the names of devices, motivations, etc., that were not included in the given categorization. Finally, although we measured loneliness and attitude towards technology, we could not see any effect of these two factors on having a smart home device. This may be in part due to having a limited number of participants who did not own a smart home device and were not positive about having one. Future work can further investigate how general attitude towards technology may influence perception of benefits of smart home devices, and how having a smart home device affects loneliness. Future work would also benefit from further studying the concerns for using smart home devices for older adults by recruiting more older adult participants, and studying concerns for using smart home systems in homes with children with participants who have small children.

8. Conclusion

Smart home devices have become more popular over the past few years and have the potential to improve quality of life. With the prominence of these technologies, it is important to understand different factors that can contribute to their success and failure, which will inform user centred design and development of these devices in the future to increase their success. To do so, we studied which smart home devices are mostly owned and how people use smart home devices, as well as what is the reason for not having a smart home device. We further studied motivations, functionalities, and concerns, which can inform design decisions. Finally, we studied if/how smart home devices can benefit people during a pandemic like COVID-19 based on users’ experiences and opinions. By expanding knowledge on how these devices are used and perceived by users or potential users, we hope that our findings can help researchers and designers to better understand users’ needs, and can inform the design of smart home technologies that can be adopted successfully by users of different ages.

Financial disclosure

This research was undertaken, in part, thanks to funding from the Canada 150 Research Chairs Program.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.