Profiles and Predictors of Smart Home Technology Adoption by Older Adults

Abstract

The commercial popularity of smart home (SH) technology has broadened the scope of aging-in-place and home health occupational therapy. The objective of this article is to examine ownership of SH technology by older adults, their readiness to adopt SH technology, and identify the client factors relating to the adoption. A survey of older adults aged 60 and above living in the community was conducted. Respondents (N = 445) who were women; in the age group of 60 to 70 years; living in a two-level home, with a body function impairment; with a fall history; and experienced in information and communication technology (ICT) were significantly likely to be “brisk adopters” of SH (p < .05). Stepwise regression model identified marital status, home security, and overall ICT ownership as the predictors of SH ownership, whereas being female, concern over home security, and perceived independence contributed to SH readiness (p < .05). Consideration of the identified client profiles, health, and personal factors will strengthen SH integration for aging-in-place.

Introduction

Smart home (SH) technology can facilitate aging-in-place for seniors by promoting safety and security, emergency response, health management, and independence in occupational routines. An increasing aging population combined with the alarming cost of long-term care relocation make SH technologies significant for older adults today. Recent advancements in information and communication technology (ICT) to support home automation and monitoring have brought SH intervention to the forefront. Although availability and affordability have significantly improved, the acceptance of SH technologies by older adults has historically been low. This article reports on a survey research to investigate the extent of older adults’ adoption of various home automation technologies as well as the factors predicting their affinity to SH.

Background

SH technology is now a popular product commodity in the consumer electronics industry. The technology’s popularity has been spurred by the integration of wireless networking, connectivity to interlink and control gadgets and appliances (termed as Internet of things), as well as to remotely monitor one’s home. The SH concept is defined as having three core characteristics: (a) sensor networks to monitor and gather information about the state of the home and its residents, (b) mechanisms that allow communication between devices to enable automation and remote access, and (c) user interfaces such as home displays, personal computers, tablets, and smart phones to enable consumers to set preferences/goals as well as to receive information and feedback (Cook, 2012; Hargreaves & Wilson, 2013). In addition, the introduction of wearable interface technologies such as smart watches and health monitors have added a new dimension to SH technology. Although home owners in general may value the comfort, convenience, and home energy management with automation (Wilson, Hargreaves, & Hauxwell-Baldwin, 2017), SH technologies have far greater implications for the aging.

A great majority of people prefer to age in their home and community (American Association of Retired Persons, 2014). Aging-in-place is now also being promoted worldwide by policy makers and health providers to save costs on institutional care (Peek et al., 2014; World Health Organization, 2007). With the impending risks of institutionalization including functional declines, physical and cognitive impairments, chronic diseases, declining social network, and low levels of physical activity (Stuck et al., 1999), the value of SH technology becomes significant with aging. Reliance on SH in the home health setting has grown as technologies continue to be widely recommended for monitoring physiological and functional outcomes, safety, social interactions, emergency detection, and cognitive and sensory assistance (Demiris et al., 2004; Liu, Stroulia, Nikolaidis, Miguel-Cruz, & Rios Rincon, 2016). More recent developments with wireless connectivity and Internet of things highlight SH technology’s capability to promote independence, telehealth, and health monitoring (Majumder et al., 2017). Chung, Demiris, and Thompson (2016) concluded that when SH technology is implemented appropriately and ethically, it has the potential to strengthen older adults’ quality of life, safety, and prospects for aging-inplace. Although the potential is noteworthy, home health professionals need to be cognizant of older adult’s context and personal factors leading to successful adoption of SH interventions.

Although valid prevalence data on ownership and usage of SH technology by older adults are yet to emerge, studies have historically highlighted a consistent level of resistance and attitudinal barriers among older adults toward SH technology (Coughlin, D’Ambrosio, Reimer, & Pratt, 2007; Demiris et al., 2004). Another precursor to technology adoption by older adults to consider is their level of readiness, defined as “the propensity to embrace and use new technologies for accomplishing goals in home life and at work” (Parasuraman, 2000, p. 308). As most older adults spend the bulk of their years in the pre–digital Internet era, their readiness to explore and use ICT is often called into question. In adoption literature, early adopters are characterized as having relatively higher awareness of innovation, wealth, information-seeking skills, and above all, the perception of the benefits of adoption (Rogers, 2003). In fact, these characteristics were evident in early adopters of SH (Wilson et al., 2017). At the other end, initial barriers to SH technology adoption by older adults are noted to be lack of awareness, limited knowledge of availability, and funding constraints (Morris et al., 2013). Concerns pertaining to usability, reliability, trust, privacy, accessibility, and affordability that were reported a decade ago (Coughlin et al., 2007) still seem to persist (Garg & Kim, 2018). There is an overwhelming perception among consumers in general that SH will increase dependence on technology, are nonessential luxuries, are intrusive, and undermine privacy (Wilson et al., 2017). In a systematic review on acceptance of SH, Peek et al. (2014) enumerated six qualitative themes associated with preimplementation of SH by older adults with a major corpus of the literature outlining the theme of concerns with technology. Along with concerns, the researchers concluded that adoption is dictated by perceived benefits and need for technology, consideration of alternatives, social influence, and personal characteristics of older adults.

As aging-in-place proponents continually elucidate the promise of SH, the assertions are yet to be supported by research on their real-world effectiveness. Evidence to support SH intervention is still emerging. Currently, most studies have small and heterogeneous samples and involve qualitative approaches (Liu et al., 2016). Although empirical evidence to support the use of SH for aging is needed to promote its role in health care (Layton, 2015), investigations into older adults’ current levels of usage and readiness for adoption are equally paramount for occupational therapists and technology providers in a home health setting. Building on the research to examine older adults’ key demographic and health determinants for adoption of ICT (Elliot, Mooney, Douthit, & Lynch, 2014; Macedo, 2017; Vroman, Arthanat, & Lysack, 2015), a parallel emphasis must be given specifically to SH technology. This study aimed at addressing two broad research questions:

• Research Question 1: To what extent has SH technology been adopted by older adults? What types of SH devices have high rates of adoption, potential for adoption, and rejection among the population?

• Research Question 2: What are the demographic and health-related factors related to the overall adoption profiles of SH technology by older adults and to what extent do the factors contribute to their SH ownership and readiness?

Health-related factors of focus were medical diagnosis; chronic disability attributed to mobility, vision, hearing, or cognitive impairments; history of falls or accidents; and perceived independence in daily routines.

Method

This descriptive study was conducted through a survey of older adults residing in the New England region of the United States. The overarching goal was to examine the status and potential as well as facilitators and barriers for them to age in place. The research protocol was reviewed and approved by the institutional review board for human participant protection at the University of New Hampshire. The data segments on home automation and the associated variables were the focus of this study.

Participants

A convenience sample of individuals 60 years and older was employed in the study. With our focus on aging-in-place, they were eligible to participate if they lived in the community alone or with family. Conceivably, older adults who lived in long-term care facilities such as assisted living and nursing homes were not considered. The chosen sample size was above 400, which is the recommended number for generalizability of findings for any population above 10,000 people for surveys that involve any categorical data and estimated margin of error of 0.05 (Bartlett, Kotrlik, & Higgins, 2001). The sample was also adequate for the multiple regression analysis employed in this study considering the recommended ratio of 10 respondents for each predictor variable (Bartlett et al., 2001).

Survey Questionnaire

A draft of the survey questionnaire was initially developed from literature on aging-in-place, and second through interviews with 10 older adults in the community. The interview participants were sampled purposively based on unique demographic characteristics, living situation, and health status. To ensure that we captured wide-ranging aging-in-place perspectives, the sample included individuals in high and low income (below the median), living alone or with family, and those with and without a chronic disability. The interview questions centered on facilitators and barriers to aging in the community including perspectives on SH technology. The interview data were analyzed by three members of the research team independently. Although the perspectives varied, data saturation was noticeable during our triangulation through common indicators and converging themes among the participants. Measurable indicators from the content analysis were added as questions on the survey. These were compiled in the areas of participation and independence in home and community activities, access and safety, use of ICT including home automation, community resources, and social support. The survey was then pilot tested with 12 older adults who first completed the questionnaire and then provided feedback and suggestions through a 2-hr focus group. Specific to this study, participants stated their needs with home automation, preferences, and dislikes; sought clarifications; and provided input on the technologies to be included in the survey. Thirteen SH commercially available devices (see Figure 1 in “Results” section) were discussed and listed in the survey and respondents needed to report on each whether they (a) “already have,” (b) “do not have but wish to have,” or (c) “do not have and do not wish to have” the device. Although the device listing was not exhaustive, the list was deemed as an adequate representation of the current SH technology market. The final survey was created following analysis of the audio-recorded data.

Figure 1.

Adopters of SH technology by demographics.

Note. SH = smart home.

Data Collection

The survey was primarily administered online through Qualtrics® survey platform. Participants were recruited through Qualtrics® Panel, a large sample pool organized by the company in various demographic and customer profiles. Potential respondents register into the panels and provide their personal and demographic information. The company then invites them to participate in surveys that match their profile and interests.

Older adults in the New England region of the United States were stratified from the sample pool. Respondents were screened out automatically if they reported their age to be below 60 years or residing in long-term care facilities. To ensure response quality, a coordinator at Qualtrics® carefully reviewed the first 50 responses for reliability and any missing sections of data. Once reliability of the panel was established, the survey was launched. A student research assistant monitored all individual responses thereafter. To check for duplication, responses were reviewed for distinct geographical location as generated from the respondent’s IP address. Comments at the end of each survey section also attested to the response validity. In addition, questions with quality filters were included at random points in the survey, in which participants were expected to respond to a question using a certain choice on the given Likert-type scale. The filters indicated that participants were paying attention to the questions and any response that deviated from the assigned choice was screened out for review and possible exclusion. Recent studies have compared the merit of online survey panels and surveys that employ traditional methods of respondent recruitment (Heen, Lieberman, & Miethe, 2014; Weinberg, Freese, & McElhattan, 2014). Although demographic variations were seen among the respondents, the quality of the data sets and findings was not significantly different in both methods.

In addition to the online version, students in our occupational therapy program administered the survey to a cohort of older adults in the community as part of a service learning assignment on aging-in-place. The survey remained active for a 2-month period from March to May 2018, and the average time for response completion on the online and hard copy versions were about 15 min and 25 min, respectively. No incentives were offered to the survey respondents through the research project, and participation was invited only based on their interest with the research topic. However, Qualtrics respondents may have received a small incentive based on the length of the survey, their specific panelist profile, and survey completion difficulty. Incentives may have included cash, airline miles, gift cards, redeemable points, sweepstake entries, and vouchers.

Data Analysis

Data relevant to this study were extracted and organized from the survey data set for analysis in IBM SPSS software. Descriptive statistics were first conducted to analyze demographics, SH technology ownership, and consumer profiles. Profiles were created based on the responses to the 13 SH devices on a 3-point ordinal scale ranging from 2 = already have, 1 = wish to have, and 0 = do not wish to have. Therefore, the total adoption score for participants ranged from 26 (for those who owned all 13 devices) to 0 (who owned no devices). Participants were then profiled into three proportionate groups—brisk adopters, emerging adopters, and slow adopters—using 33.3 percentile distribution in the score. Cross tabulations along with chi-square analysis of the three adopter profiles were conducted against key variables of demographics, living situation, ICT, and health. Studies in the past have used a similar analytical framework to examine profiles of older adults who adopted ICT (Vroman et al., 2015).

To examine the key predictors to SH adoption, the outcomes of ownership and overall readiness were computed in continuous scale for each respondent. The ownership score was the total number of devices owned from 0 to 13, and overall readiness score was calculated by adding the devices owned and wished to have, and deducting the number of devices that they did not wish to have from the sum. The readiness score, therefore, ranged from +13 to −13. Doing so negated any bias or relative importance to the devices in the analysis and accounted current ownership and interest, as well as reluctance to SH technology. For the prediction analysis, bivariate correlation coefficients were first calculated individually between key predictor variables and the outcome variables of ownership and overall readiness. The predictor variables were chosen based on their relevance as well as association with SH adoption as examined in the cross tabulations. We then conducted a stepwise multiple regression analysis from the bivariate analysis with predictors with p value less than .1 to generate the model that best contributed to the ownership of, and readiness for, SH technology. Multicollinearity was verified and any predictor variable with tolerance less than .1 and variance inflation factor above 5 was removed from the model.

Results

In all, 445 older adults completed the survey—416 respondents from the online Qualtrics panel and 29 interviewed by students. The average age of the sample was 70.7 years (SD = 5.3 years, range = 60–95 years) with the majority (68%) being females. Participant demographics including residential information are displayed in Table 1. Fifty-four percent were married and about 80% had some college education and higher. The majority (78%) were retired and the household income was well distributed with about 20% reporting income more than US$60,000, around the national median household income. About 55% lived with a spouse or partner, whereas 35% resided alone. The geographical locations were evenly distributed with participants reporting their residence in cities, suburbs, towns, and rural small towns. Sixty-one percent lived in a two-level home. On the question of home security (please rate how you feel about the security of your home) on a 5-point Likert-type scale, the majority felt their home to be safe (30%) to very safe (64%).

Table 1.

Participant Demographics.

Characteristics	%	Frequency
Gender (N = 444)
Males	31.7	141
Females	68.1	303
Marital status (N = 442)
Single	18.7	83
Married	53.9	240
Divorced	20.9	93
Separated	1.6	7
With partner	4.3	19
Education (N = 441)
Below high school	0.4	2
Completed high school	18.4	82
Some college	22.2	99
Associate degree/diploma	10.3	46
Bachelor’s degree	25.2	112
Master’s degree or higher	22.5	100
Income (N = 438)
Below US$15,000	5.6	25
US$15,000–US$30,000	17.1	76
US$30,000–US$45,000	16.9	75
US$45,000–US$60,000	16.6	74
US$60,000–US$75,000	10.6	47
US$75,000–US$90,000	9.9	44
Above US$90,000	21.8	97
Employment (N = 440)
Full time	7.6	34
Part time	8.1	36
Self-employed	4.7	21
Unemployed and seeking job	1.6	7
Retired	76.9	342
Ethnicity (N = 444)
White (Caucasian)	95.7	425
African American	1.1	5
Hispanic or Latino	0.7	3
Asian	0.9	4
Native Indian	0.4	2
Other	1.1	5
Place of living (N = 444)
City	20.7	92
Town	23.6	105
Suburb	28.3	126
Rural or small town	27.2	121
Living situation (N = 444)
With spouse or partner	54.8	244
Alone	35.3	157
With family	9.7	43
Home design (N = 443)
One level	38.9	173
Two or more levels	60.7	270

Seventy-three percent of the respondents reported having a medical condition, many of which were chronic, yet manageable, such as hypertension and diabetes. Fifty-three percent reported at least one body function impairment—23% in mobility, 22.7% in postural balance, 29% in vision, 19% in hearing, and 10.4% in cognition. A small, yet notable, proportion (14%) of older adults had experienced at least one injurious fall or accident at home. When asked to rate their independence in managing daily personal routines on a 5-point Likert-type scale, 93% reported being independent to very independent. With respect to ownership of ICT, a very high proportion of older adults had a computer (98.4%), and had access to Internet (98.9%), cell phone (94.6%), and cable television (90.1%). Looking at devices to facilitate SH wireless connectivity and remote monitoring, majority of respondents used a smart phone (81%) and tablet (74.5%), whereas about half (53%) reported that they owned a smart watch.

Profiles of SH Technology Adoption

Table 2 highlights the percentage of older adult respondents (N = 445) in rank order of SH device ownership along with wish for ownership, and reluctance for ownership. For this preliminary analysis, respondents were classified accordingly as current adopters, potential adopters and nonadopters for each of the 13 devices. The devices with high rates of ownership were carbon monoxide alarms, thermostats, motion sensing lights, and backup power generator. Interest for the top five devices ranked from water leak detectors, auto shutoff on stoves, backup generator, and motion-sensing lights, whereas the least desired devices were remote home monitoring, control of lights and appliances via smart phone, motion-activated camera, voice-activated assistant, and emergency alerting system. Examining the median distribution, respondents, on average, owned three of the listed 13 devices (M = 3.2, SD = 2.01), wished for an additional three (M = 3.3, SD = 3.11), and expressed no interest in six devices (M = 6.3, SD = 3.3). The median for the total adoption score was 10 (M = 9.8, SD = 4.46) with a range of 0 to 23. Using 33.3 percentile ranges, three proportionate groups were derived. Slow adopters (30%) had a total adoption score ranging from 0 to 8. Emerging adopters’ (33%) score ranged from 9 to 12 and brisk adopters (37%) scored at 13 or higher. Note that the groups could not be derived in three exact proportions due to the lack of decimal points in the overall adoption score.

Table 2.

Ownership and Interest in Common SH Devices (N = 445) with top five devices in each category in bold.

SH devices	Current adopters (%)	Potential adopters (%)	Nonadopters (%)
Carbon monoxide alarm	81.5	12.2	6.3
Manually programmable thermostat	47.5	20.8	31.7
Autoset thermostat	43.8	17.7	38.5
Motion sensor lights	35.8	27.3	36.9
Backup generator	27.9	36.5	35.6
Home security system	19.5	23.8	56.8
Voice-activated assistant	19.3	17.2	63.5
Emergency alert system	17.2	23.4	59.4
Water leak detector	9.9	41.8	47.4
Motion-activated camera	7.7	25.2	67.0
Auto shutoff stove	7.2	38.0	54.8
SH control	5.4	25.4	69.2
Remote monitoring	5.0	23.8	71.3

Note. SH = smart home.

Cross tabulations and Pearson chi-square analysis of the adopters in two age categories (see Figure 2) show that individuals 60 to 70 years old are significantly more likely (χ²(1, N = 443) = 5.6, p < .05) to be brisk SH adopters (41.6%) than those 71 years and above (30.6%). In terms of gender, more women were brisk adopters (39.1%) with a significant association as opposed to men who were comparatively slow adopters (40.4%; χ²(1, N = 443) = 10.8, p < .05). There were no patterns or significant associations seen with the older adults’ marital status, education, income, or place of living. However, with the design of the home, older adults living in two-level homes were significantly more likely to be brisk adopters (40.7%) than those in one-level homes (30.8%; χ²(1, N = 443) = 6.4, p < .05).

Figure 2.

Adopters of SH technology by home design.

Note. SH = smart home.

With respect to associations with health, individuals with and without medical diagnosis have an almost equal likelihood to be brisk (38% and 34.2%), emerging (32.1% and 35.8%), and slow adopters (29.9% and 30%, respectively). However, there were a significantly higher number of older adults with body function impairments including problems with mobility (χ²(1, N = 443) = 6.3, p < .05), balance (χ²(1, N = 442) = 7.6, p < .05), and hearing (χ²(1, N = 441) = 4.1, p < .05) who were more likely to be brisk and emerging adopters than slow adopters, as seen in Figure 3. No similar patterns were evident in those with sensory and cognitive impairments. Also, older adults who experienced an injurious fall or accident (14%) tended be brisk adopters (42.6%) than slow adopters (19.7%) of SH technology although the association was not statistically significant (χ²(1, N = 443) = 3.46, p = .17). Figure 4 shows the link that older adults who owned ICT devices that provide portability, wireless connectivity, and remote monitoring are significantly likely to be SH adopters. Majority of individuals who owned smart phone, 40%, were brisk adopters compared with the 23.8% who did not own one (χ²(1, N = 444) = 7.8, p < .05). The same pattern with brisk adopters was noticeable with those who used a tablet, 42.6% and 20.4% (χ²(1, N = 444) = 18.4, p < .05), or a smart watch, 44.3% and 28.7% (χ²(1, N = 444) = 11.5, p < .05).

Figure 3.

Adopters of SH technology by functional impairments and safety.

Note. SH = smart home.

Figure 4.

Adopters of SH technology by ICT ownership.

Note. SH = smart home; ICT = information and communication technology.

Predictors of SH Adoption

Tables 3 and 4 list findings from the bivariate and stepwise regression analysis for the outcomes of SH ownership and readiness, respectively. For ownership, the significant predictors were being married; income above US$60,000; living with a spouse or partner; home’s perceived security; any body function impairment with mobility, balance, vision, hearing, or cognition; and total ownership of ICT. Stepwise regression from the bivariate list (p < .1) indicated being married, home security, and ICT ownership as the significant predictors in the model (F(3, 434) = 5, p < .05, R² = .08).

Table 3.

Predictors of SH Ownership (with significant predictors (p<0.05) highlighted).

Variables	Bivariate model					Stepwise regression model
	B	SE B	ß	T	p	B	SE B	ß	t	p
Age	0.00	0.02	0.01	0.23	.82
Gender—female	0.31	0.21	0.07	1.49	.14
Married	0.71	0.19	0.18	3.73	.00	0.6	0.19	0.15	3.1	.002
College education and above	−0.39	0.24	−0.08	−1.62	.11
Income > 60k	0.47	0.19	0.12	2.46	.01
Living with spouse/partner	0.52	0.19	0.13	2.71	.01
Home security	0.54	0.14	0.18	3.94	.00	0.42	0.13	0.15	3.16	.002
Medical diagnosis	0.38	0.21	0.08	1.77	.08
Body function impairment	−0.54	0.19	−0.13	−2.85	.00
Number of falls	0.00	0.08	0.00	−0.04	.97
Independence in daily routines	−0.02	0.14	−0.01	−0.17	.87
ICT ownership	0.18	0.05	0.18	3.87	.00	0.14	.045	0.14	3.02	.003

Note. SH = smart home; ICT = information and communication technology.

Table 4.

Predictors of SH Readiness (with significant predictors (p<0.05) highlighted).

Variables	Bivariate model					Stepwise regression model
	B	SE B	ß	T	p	B	SE B	ß	t	p
Age	−0.12	0.06	−0.10	−2.08	.04
Gender—female	1.83	0.67	0.13	2.73	.01	1.25	0.45	0.13	2.8	.05
Married	−0.08	0.63	−0.01	−0.12	.90
College education and above	0.72	0.79	0.04	0.92	.36
Income	−0.27	0.64	−0.02	−0.42	.67
Living with spouse/partner	−0.03	0.63	0.00	−0.05	.96
Home security	−1.21	0.46	−0.13	−2.64	.01
Medical diagnosis	−0.31	0.71	−0.02	−0.44	.66
Body function impairment	1.14	0.63	0.09	1.81	.07
Number of falls	0.53	0.26	0.10	2.04	.04
Independence in routines	−1.15	0.46	−0.12	−2.51	.01	−0.63	0.30	−0.096	−2.06	.04
ICT ownership	0.47	0.15	0.14	3.04	.00	0.45	0.10	0.19	4.25	.00

Note. SH = smart home; ICT = information and communication technology.

The average readiness score for SH technology was 0.3 (SD = 6.6) with a median of 1 and range from 13 to −13. Increasing age, higher perceived independence in daily occupational routines, and home security were negative predictors (p < .05) of readiness. At the same time, being female, number of falls, and overall ownership of ICT contributed to readiness for SH technology (p < .05). Stepwise regression pointed out being female, independence in routines, and overall ICT ownership to be the significant predictors contributing to the model (F(3, 437) = 9.4, p < .05, R² = .06).

Discussion

The study findings reveal the extent of SH technology adoption among older adults, their ownership and affinity toward specific types of SH devices, their consumer profile, and factors attributed to current and future adoption. Overall, the adoption rate of some of the popular SH devices was considerably low as reflected in past findings (Liu et al., 2016). However, certain devices had comparatively higher rates of adoption such as the carbon monoxide alarm, programmable and autoset thermostat. Safety seemed to be the clear priority in older adults’ preference for SH, with a major portion of the sample indicating interest in devices such as the water leak detector, auto shutoff for stove, and backup generator. Interestingly, the survey findings reflected the diffusion of innovation theory (Rogers, 2003) and supported the prevailing notion of older adults being late adopters of SH technology (Wilson et al., 2017). Adoption seemed chronological in the list, with most of the recent innovations such as voice-activated assistants, motion-activated camera, smart light control, and remote home monitoring being the least preferred by the older adults in our sample. It will be interesting to examine how the adoption pattern shifts with future innovations in SH technology.

Demographic Factors With SH Adoption

Our data indicate women in their baby boomer age of 60 to 70 years (more than men) to be emerging consumers of SH. Although ownership of SH technology was not predicted by age and gender, increasing age and being male significantly lessened the readiness to adopt it in the future. This finding was unanticipated as women have traditionally been slow to adopt ICT in general compared with men (Hargittai, 2010). Nevertheless, gender differences in adoption of specific types of ICT are not uncommon and that may be the case with SH technology. For example, whereas men have predisposition toward entertainment-based ICTs, women have been found have a relative preference for social interaction–based ICTs than men (Büchi, Just, & Latzer, 2016). Contrary to our expectation, older adults at all levels of education and income have an equal likelihood of adopting SH technology. A plausible explanation to this finding may be that most of the SH devices that were owned were commercially well known and affordable across all demographics of older adults. Although the subsequent regression analysis did show income more than US$60,000 as a significant predictor of ownership, the overall readiness for SH was not predicted by income. Being married and living with a spouse or partner contributed significantly to current ownership of SH as indicated in previous studies on adoption of ICT (Vroman et al., 2015). A novel finding from the study was that older adults residing in two-level homes were far more likely to realize the need and adopt SH technology sooner than those living in a one-level home with space possibly being an influencing factor.

Health and SH Adoption

As theorized by Davis (1989) and later highlighted in studies, the realization of technology’s benefit and usefulness serves as a vital catalyst to its adoption by older adults (Lee & Coughlin, 2015; Yusif, Soar, & Hafeez-Baig, 2016). Older adults with a body function impairment and those with a history of a major fall or accident were potential adopters of SH technology possibly due to their evolving concern for safety, security, and independence. Analysis of our data reveals the benefits of SH technology in promoting a sense of security. There was a positive correlation between SH ownership and perceived home security, although concern for security significantly increased readiness to adopt newer SH devices. To highlight another predictor, the potential of SH technology to support participation became evident in the study findings. Higher perceived independence in managing daily routines was negatively associated with readiness, and we may deduce inversely that declining independence may prompt the need for SH technology.

Role of ICT

Our study sample had a notably higher degree of ICT ownership (including computer and Internet) compared with the figures reported in nationwide survey studies (Pew Research Center, 2017). The findings confirm that older adults with exposure and access to the more recent and advanced ICTs such as tablets, smart phones, and activity tracking watches are significantly more likely to seek out and adopt SH technology. As a common thread, all respondents who used remote home monitoring, the 92% who owned a voice-activated assistant, and the 74% who had installed a motion-activated camera, all owned a smart phone. Overall, ICT ownership (especially with the devices that were designed around portability, connectivity, and network integration within the home) paved way for SH adoption. The finding is in sync with product design literature that technology experience and domain knowledge positively influence consumer skill and comfort levels with newer technologies (Jordan, 1998). Also, as per innovation diffusion theory (Rogers, 2003), ICT may be serving as the social channel for older adults to facilitate preimplementation knowledge of SH.

Clinical Implications

Occupational therapy has been at the forefront for fostering participation for older adults in the context of home and community. The profession has pioneered advances and contributed to a deep knowledge base in home assessments and environmental modifications. SH intervention is an evolving subdomain, and it is important for the profession to keep pace with rapid advancements in home automation, to ensure interventions are delivered using core practice principles and evidence-based best practices. The challenge for now is that quantity and quality of evidence are low (Liu et al., 2016). Formal approaches are necessary to examine the client factors that may lead to successful adoption or abandonment of the technology. Preintervention evaluations need to elaborately gather the client’s demographic factors and health history including body structure and functions, home safety, and independence. These premorbid factors may be an inherent advantage for some older clients. However, it is also important to acknowledge that the predictors identified in our models only contributed to 8% and 6% of the variance (R²) with SH ownership and overall readiness, respectively. Regardless of the client factors, a major goal of the intervention must also be in reducing technology anxiety and facilitating motivation, perceived benefit, and trust of older clients to embrace SH technology. For instance, a recent end user study found that the behavioral traits of performance expectancy, effort expectancy, expert advice, and trust together contributed about 81% of the variance in an older adult’s intent to use SH technology for meeting health care needs (Pal, Funilkul, Charoenkitkarn, & Kanthamanon, 2018).

Study Limitations

The study sample was limited to the New England region of the United States and had unique demographic characteristics such as income and ethnicity. Therefore, the findings in entirety may not generalize nationally or globally to the aging population. The study used two distinct data sources, with the vast portion of data collected through the online panel. Nonetheless, similar to past research that compared the merit of the two sources (Heen et al., 2014; Weinberg et al., 2014), there were no noticeable differences found in key measures of SH ownership, readiness, and overall adoption between the two sample cohorts. The analytical framework used to define the profiles and outcomes of ownership and readiness were arbitrary to the study, and future research on SH adoption may need to examine whether the findings are consistent.

Conclusion

Innovations in home automation technology are surging. However, adoption and interest in SH technology are relatively low among older adults. Current levels of ownership and readiness vary vastly by the type of technology, demographic segments, functional status, and home safety. The findings from this study are expected to contribute to the better integration of SH interventions in occupational therapy home health setting. Future research may examine the explored factors in an intervention context with smaller cohorts of clients to further elucidate barriers and best practices for practitioners to promote SH technology adoption.