I Want It Anyway Consumer Perceptions of Smart Home Devices
I Want It Anyway Consumer Perceptions of Smart Home Devices
I Want It Anyway Consumer Perceptions of Smart Home Devices
To cite this article: Xuequn Wang, Tanya Jane McGill & Jane E. Klobas (2020) I Want It Anyway:
Consumer Perceptions of Smart Home Devices, Journal of Computer Information Systems,
60:5, 437-447, DOI: 10.1080/08874417.2018.1528486
ABSTRACT KEYWORDS
Smart home devices form a significant part of the Internet of Things market and can provide benefits Internet of things; smart
such as convenience and energy efficiency. They also have potential privacy and security risks as they home devices; adoption
collect information constantly. In order to examine how benefit and risk factors influence individuals’ intention; net valance model
intentions to adopt smart home devices, we developed a net valence model that integrates both
positive factors and risk factors. The model was tested using data collected using an online question-
naire. The results show that individuals tend to ignore the potential risks and focus more on potential
benefits resulting from using smart home devices. Performance expectancy and compatibility were
found to be positively related to perceived benefits. However, neither effort expectancy nor image were.
Among the proposed dimensions of risk, only privacy risk, performance risk, and time risk significantly
influenced perceived risk. Security risk and financial risk did not influence it.
CONTACT Xuequn Wang [email protected] Murdoch University, 90 South St., Perth 6150, Australia
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/ucis.
© 2018 International Association for Computer Information Systems
438 X. WANG ET AL.
home devices,14 while in others overall perceived benefits and/ investigate the effect of individuals’ perceptions of opposing
or attitude are included as mediating their effect.16,19,21,22 benefit (positive valence) and risk (negative valence) factors on
Less attention has been paid to the risk factors that may their decisions.27–29 The basic proposition of these NVMs is that
have a negative influence on adoption, and what research individuals are more likely to engage in a certain behavior (e.g.,
there has been has focused primarily on privacy risk, with using smart home devices) if the perceived benefits of the beha-
studies finding that factors that have been labelled as per- vior outweigh the perceived risk. These NVMs have been used to
ceived privacy risk,13,22 concern for information privacy,21 understand intention to continue using social media,27 health
perceived sacrifice,16 and security/privacy risk19 play either information seeking/sharing,28 and intention to participate in
direct or indirect roles in reducing intentions to adopt smart Bring-Your-Own-Device (BYOD) programs.29
home devices or services. Hsu and Lin21 considered four In the context of technology adoption, the perceived ben-
kinds of privacy issues: collection of data, unauthorized sec- efits of the technology of interest are assigned a positive
ondary use, improper access, and errors (in data). However, valence.30 Since not all perceived benefits can be quantified
no other risk factors were considered beyond privacy risk. The objectively, individuals are assumed to rely on their own
study by Dong, et al.13 had an even narrower conceptualiza- subjective evaluations when they make technology and use
tion of privacy risk, measuring it with three items that related decisions. Based on UTAUT17 and attributes of technology
only to stealing of personal information. innovations that Moore and Benbasat31 derived from Rogers’
In addition to privacy risk, Yang, et al.19 also considered model of adoption and diffusion, we propose performance
physical risk and trust of service providers. They found that expectancy (perceived usefulness), effort expectancy (ease of
potential adopters were not concerned about physical risk, but use), compatibility, and image as factors relevant to the per-
that trust in the service provider had a positive effect on ceived benefits of smart home usage. Whether individuals are
adoption. Gao and Bai14 also included the role of trust, rather positive or negative about these potential benefits, the concept
than risk, but found no influence on adoption. Their concep- of perceived benefits retains its positive connotation in the
tualization of trust did not explicitly address privacy and NVM. For example, when perceived usefulness is low, the
security risk – rather confidence in the provider and reliability technology is perceived to be difficult to use, incompatible
of the information from the system. with current technology and activities, or use of the technol-
In a study on adoption of smart home services, Hsu and ogy contributes to a negative rather than a positive personal
Lin22 considered the role of perceived fees of services, but image, the value of perceived benefits is low, but the concept
found that financial cost did not influence adoption inten- of perceived benefits retains its positive valence.
tions. Kim, et al.16 also included perceived fees in their con- Perceived risk is the subjective expectation of a possible
ceptualization of perceived sacrifice but did not have a loss.32 NVMs of technology adoption assign a negative
separate hypothesis associated with it. So although they valence to perceived risk, which is defined in terms of indivi-
found that perceived sacrifice did have a negative impact on duals’ expectation of negative consequences from adoption or
perceived value, the relative influence of perceived fee relative use of a technology.9,28 Following refs.,33,34 individuals’ per-
to the other risk they considered (technicality) was not deter- ceptions of the risk associated with a behavior are expected to
mined. Given the limited research that has been done on influence their intention to engage in that behavior.
smart home device risk and the apparently inconsistent results Following perceived risk theorists34,35 and previous litera-
of research that has been reported, further research is needed ture on technology adoption8,9 we propose privacy risk, secur-
to understand the nature of different risk perceptions and ity risk, performance risk, time risk, and financial risk as
their relative roles in adoption. perceived risk factors relevant to smart home device adoption.
Performance
Expectancy
H2a
Effort H2b
Expectancy
H2c Perceived
Compatibility
Benefits
H1
Image H2d
Intention to
Privacy Risk Use Smart
H4a Home Devices
H4b
Security Risk
Performance H4c H3
Perceived Risk
Risk
H4d
Time Risk
Figure 1. A Net Valence Model (NVM) of intention to use smart home devices.
of use) has a positive association with intention to use IoT. Privacy risk refers to the “potential loss of control over
Following the literature, we argue that individuals are more personal information.”9 Since individuals need to control
likely to choose those devices that are easier to use and more smart home devices through other devices (e.g., smartphones)
user friendly because of the influence of effort expectancy on over the Internet, using these devices creates the potential for
perceived benefits. Following UTAUT, we propose that: personal information to be disclosed to providers of devices
without individuals’ consent and knowledge. Individuals may
H2b: Effort expectancy is positively related to perceived benefits hence feel concerned about their privacy when using smart
of using smart home devices. home devices. Previous research has found that perceived
privacy risk has a negative effect on individuals’ attitude
Compatibility has been defined as “the degree to which an toward using IoT services21 and smart home services.19
innovation is perceived as being consistent with the existing Thus, we propose that:
values, needs, and past experiences of an individual.”31 Hsu
and Lin21 found that perceived compatibility is positively H4a: Privacy risk is positively related to perceived risk of using
related to the perceived benefits of IoT services, and Park, smart home devices.
et al.37 found that it is positively related to intention to use
smart home services. Therefore, we proposed that those who Smart home devices are connected with other devices such
perceive smart home devices as compatible with the way they as control panels, smartphones, and computers. Such inter-
live are more likely to perceive the use of smart home devices connected systems result in the potential for security
as beneficial and hypothesize that: breaches.5,6 In a study on customer acceptance of RFID tech-
nology, Müller-Seitz, et al.43 found that security concerns
H2c: Compatibility is positively related to perceived benefits of negatively impacted on acceptance of the technology and
using smart home devices. similarly Lee8 found that security risk had a negative influence
on both attitude to internet banking and intention to use it.
Image is defined as “the degree to which use of an innova- We, therefore, propose that if individuals believe that using
tion is perceived to enhance one’s image or status in one’s smart home devices can cause security breaches that harm
social system,”31 and has been found to play a role in adoption their daily lives, this will influence their overall risk perception
of electric cars,38 smartphone brand loyalty,39 and in under- and we hypothesize that:
taking home automation projects.40 We propose that potential
users may perceive that using smart home devices will H4b: Security risk is positively related to perceived risk of using
enhance their image and that this will increase their benefit smart home devices.
perceptions. Therefore, we propose that:
Performance risk reflects “how well the product will
H2d: Image is positively related to perceived benefits of using perform relative to expectations.”41 Smart home devices
smart home devices. are intended to enhance the convenience of consumers’
daily lives. For example, individuals expect that monitor-
ing systems will allow them to monitor the condition of
their houses when they are traveling. However, it is pos-
Perceived risk of using smart home devices
sible that selected devices may not meet individuals’
Perceived risk has been defined as “the amount that would be requirements. For example, Wilson, et al.10 identified
lost (i.e., that which is at stake) if the consequences of an act that potential consumers are concerned with potential
were not favorable, and the individual’s subjective feeling of performance risk such as disruption to daily routines,
certainty that the consequences will be unfavorable,”34 and and believe that there is a need for devices to be reliable,
perceived risk has been shown to influence adoption of innova- be made by credible manufacturers, and come with per-
tions such as online banking41 and mobile payment services.42 formance warranties. The impact of performance risk on
In the context of smart home devices, we define perceived risk overall risk perceptions has not previously been studied,
as individuals’ beliefs about the potential negative outcomes however Park, et al.37 reported that 11% of their intervie-
caused by using smart home devices and hypothesize that: wees identified system reliability as a relevant factor in
determining adoption of smart home services, and
H3: Perceived risk of using smart home devices is negatively included it in their model, and found that it has a weak
related to intention to use smart home devices. impact on perceived usefulness. Yang, et al.19 investigated
the role of potential physical risk from smart home
Perceived risk has been proposed to be a multi-dimen- devices, such as side effects due to malfunctioning or
sional construct,33 and the consumer marketing literature34 misuse, and they found that it did not affect users’ atti-
has proposed risk dimensions that are relevant in different tudes towards them, but did not specifically look at per-
contexts of technology adoption and have been investigated in formance. In this study, consistent with Featherman and
technology adoption studies e.g.,9,29. Based on this relevant Pavlou,9 we propose that:
literature, we propose that the risk associated with smart
home devices include: privacy, security, performance, time, H4c: Performance risk is positively related to perceived risk of
and financial risk. using smart home devices.
JOURNAL OF COMPUTER INFORMATION SYSTEMS 441
cant. Therefore, CMB was unlikely to be a serious concern for .90),50 (Table 5). Therefore, our measures demonstrated good
our study. psychometric properties.
Our model was tested with PLS, and SmartPLS was used.48 We then tested the structural paths in our model
The bootstrap resampling method (using 1,000 samples) was (Figure 2). As hypothesized, perceived benefits was positively
applied to determine the significance of the paths. We first related to intention to use smart home devices (ß = .72,
evaluated the measurement model. As shown in Table 3, each p < .001); thus H1 was supported. Among its antecedents,
item loaded significantly on its respective construct and none performance expectancy (ß = .72, p < .001) and compatibility
of the loadings were below .50.49 In addition, the Cronbach’s (ß = .72, p < .001) were found to be positively related to
Alphas and composite reliabilities (CRs) of all constructs were perceived benefits, supporting H2a and H2c, respectively.
over .70, and the average variance extracted (AVE) was over However, neither effort expectancy nor image influenced per-
.50 for all constructs (Table 3). Therefore, convergent validity ceived benefits, so H2b and H2d were not supported.
was supported. Discriminant validity was also confirmed by Surprisingly, perceived risk was not significantly related to
ensuring that for each construct, the square root of its AVE intention to use smart home devices (ß = -.04, p > .05); H3
exceeded all correlations between that construct and any other was therefore not supported. Among its dimensions, privacy
construct (Table 4). We also assessed Heterotrait-Monotrait risk (ß = .35, p < .001), performance risk (ß = .18, p < .05), and
Ratio of Correlations, which were also acceptable (i.e., below time risk (ß = .21, p < .01) were found to contribute to
perceived risk; therefore H4a, H4c, and H4d were supported.
Neither security risk nor financial risk contributed to perceive
Table 3. Descriptive statistics for items.
risk, thus H4b and H4d were not supported. Our model can
Construct Item Mean SD Loading ɑ CR AVE explain 72.7% of the variance in perceived benefits, and 45.1%
Intention to use smart home Int1 4.70 1.63 .94 .95 .96 .87 of the variance in perceived risk. The model accounted for
devices Int2 4.32 1.65 .90
Int3 4.61 1.71 .96 53.5% of the variance in intention to use smart home devices.
Int4 4.84 1.69 .94
Perceived benefits PB1 5.09 1.38 .95 .90 .95 .91
PB2 5.22 1.24 .95
Performance expectancy PE1 5.29 1.26 .89 .92 .87 .80 Discussion
PE2 5.18 1.31 .92
PE3 5.08 1.46 .87 This study aimed to examine how various factors influence
Effort expectancy EE1 4.89 1.35 .83 .91 .94 .74 individuals’ intention to adopt smart home devices. We devel-
EE2 4.93 1.35 76
EE3 5.00 1.45 .91 oped a NVM to identify relevant positive and negative factors.
EE4 5.03 1.44 .91 The model was tested with a US sample, and the results
EE5 4.86 1.38 .89
Compatibility COM1 4.56 1.60 .94 .93 .96 .88
provide partial support of our model. Our study makes
COM2 4.59 1.63 .95 important theoretical and practical implications.
COM3 4.59 1.66 .94
Image Image1 4.17 1.73 .87 .88 .92 .73
Image2 4.29 1.59 .85
Image3 4.46 1.46 .88 Implications for theory
Image4 4.32 1.72 .82
Perceived risk PR1 4.37 1.53 .93 .76 .89 .80 This study has several important theoretical implications.
PR2 4.30 1.54 .86 First, our study uses a NVM to examine individuals’ intention
Privacy risk PriR1 4.65 1.48 .96 .91 .90 .95
PriR2 4.67 1.49 .95 to use smart home devices. Our results show that perceived
Security risk SR1 4.67 1.41 .93 .81 .91 .84 benefits have a strong effect on adoption intention, while
SR2 5.13 1.39 .90
Performance risk PerfR1 4.30 1.46 .90 .91 .94 .79 perceived risk does not have a significant impact. These
PerfR2 4.50 1.46 .92 results show that individuals tend to ignore the potential
PerfR3 4.63 1.44 .88 risk and focus more on the potential benefits resulting from
PerfR4 4.34 1.42 .86
Time risk TR1 3.97 1.57 .90 .89 .93 .82 using smart home devices. Therefore, the theoretical model
TR2 3.85 1.623 92 developed in this study helps clarify the process of smart
TR3 4.21 1.50 .90
Financial risk FR1 5.26 1.37 .79 .83 .90 .74 home device adoption.
FR2 4.84 1.47 .89 Second, our study identifies relevant positive and negative
FR3 4.36 1.46 .90 factors based on previous relevant literature and clarifies how
Table 4. Correlation between constructs and square root of AVEs (on Diagonal).
1 2 3 4 5 6 7 8 9 10 11 12
1 Intention to use smart home devices .93
2 Perceived benefits .73 .95
3 Performance expectancy .69 .82 .89
4 Effort expectancy .61 .64 .66 .86
5 Compatibility .79 .77 .76 .71 .94
6 Image .46 .46 .49 .47 .53 .86
7 Perceived risk −.23 −.26 −.31 −.21 −.28 −.04 .90
8 Privacy risk −.15 −.19 −.22 −.14 −.24 −.08 .60 .95
9 Security risk −.15 −.19 −.22 −.13 −.22 −.08 .57 .78 .92
10 Performance risk −.17 −.22 −.30 −.30 −.28 −.14 .51 .48 .55 .89
11 Time risk −.23 −.28 −.31 −.43 −.31 −.09 .48 .41 .45 .64 .91
12 Financial risk −.08 −.07 −.13 −.25 −.17 −.06 .35 .39 .44 .62 .61 .86
JOURNAL OF COMPUTER INFORMATION SYSTEMS 443
Table 5. Heterotrait-monotrait ratio of correlations. domain,13,22 and compatibility has previously been identified
1 2 3 4 5 6 7 8 9 10 11 12 as relevant in determining the perceived benefits of smart
1 Intention to use home devices.37
smart home Effort expectancy was not, however, significantly related to
devices
2 Perceived benefits 77 perceived benefits, despite the fact that Dong, et al.13 and
3 Performance .74 .88 Liew, et al.36 found that it had a direct effect on intention to
expectancy
4 Effort expectancy .62 .63 .71 use IOT devices. The lack of relationship may be because, as
5 Compatibility .84 .80 .83 .75 individuals increasingly use smartphones and other technol-
6 Image .44 .47 .54 .50 .53 ogy, they assume that using smart home devices will be
7 Perceived risk .23 .25 .33 .20 .27 .07
8 Privacy risk .15 .14 .18 .09 .22 .07 .69 straightforward and hence the effort needed to use these
9 Security risk .14 .15 .20 .09 .20 .08 .69 .89 devices becomes a less important issue. This is consistent
10 Performance risk .16 .18 .33 .29 .27 .16 .54 .46 .58
11 Time risk .27 .32 .38 .50 .36 .15 .51 .39 .49 .67 with previous technology adoption literature17 in that effort
12 Financial risk .12 .10 .19 .27 .23 .12 .41 .43 .54 .70 .67 expectancy is less important in later stages of technology
adoption. However, smart home technologies are relatively
new, and several studies have suggested that rather than
these factors influence individuals’ risk and benefit percep- being easy to use once adopted, consumers often struggle
tions associated with smart home device adoption. To the best with them.44,51
of our knowledge, previous research has not studied such a Image also did not have a significant influence on perceived
full range of potential perceived benefit and risk factors for benefits. It is possible that as individuals increasingly use IoT
smart home device use. Of the four potential antecedents of devices, adopting smart devices in the home, where they are less
perceived benefits identified based on UTAUT17 and attri- visible than other consumer IoT devices such as fitness devices, is
butes of innovations from Moore and Benbasat,31 perfor- not perceived as a major contributor to overall perceived benefits.
mance expectancy and compatibility were found to be Among the proposed five dimensions of risk, only privacy risk,
positively related to perceived benefits. These results show performance risk, and time risk significantly influenced perceived
that when considering the potential benefits related to smart risk. These results show that individuals feel concerned about their
home devices, individuals focus on the extent to which the privacy if they were to adopt smart home devices and do have an
devices will improve their performance of household tasks awareness that these devices may take much time to set up and
and are consistent with their needs. The role of performance successfully operate. The effect of performance risk is consistent
expectancy is consistent with the findings in the broader IoT with that of performance expectancy in that individuals care about
Performance
Expectancy
.55***
Effort .05
Expectancy
R2 = .73
.32*** Perceived
Compatibility
Benefits
.72***
Image -.00
R2 = .54
Intention to
Privacy Risk Use Smart
.35*** Home Devices
.13
Security Risk
R2 = .45
potential performance risk such as disruption to daily routines and Although our results show that perceived risk is not signifi-
want the devices to come with performance warranties.10 cantly related to smart home devices adoption, consumers should
Security risk and financial risk did not have significant effects. It be aware that using smart home devices is associated with various
is possible that as smart home technology is becoming more types of risk. In other words, individuals should take these different
mature, individuals perceive smart home devices as less likely to types of risk into consideration when deciding whether to use
be hacked, and therefore security does not play an important role smart home devices. Therefore, providers of smart home devices
in their risk perceptions, despite the GfK11 finding that security should invest effort in lowering the risk associated with using the
was the third most common perceived risk of smart home devices. devices. Our results can help providers to decide which risk to
An alternative explanation may be that some consumers have low focus on. For example, as our results show that privacy risk,
awareness of the potential security risk and their implication. performance risk and time risk contribute to individuals’ overall
Future research should explore awareness and knowledge of spe- risk perception, to protect individuals’ privacy, providers should
cific security risk. The financial cost of smart home devices also establish privacy policies and inform users how collected informa-
seems to a less important issue in determining overall risk percep- tion (if there is any) will be used. Providers should also focus on
tions. Whilst Park, et al.37 found that perceived cost had a negative improving their devices so that they do not need much time to set
relationship with intention to use smart home devices, the rela- up and learn to use, thus reducing time risk perceptions.
tionship was very weak and perceived cost was not considered as a
component of overall perceived risk.
Lastly, our results add to those of previous studies applying a Limitations and opportunities for future studies
NVM. Performance expectancy, compatibility and privacy risk Our study has the limitation that the sample was recruited through
appear to have consistent effects across different technology con- a survey company. Although it includes participants from a variety
texts, for example refs.,27,28 whereas the roles of image, perfor- of backgrounds, our sample could still be biased as participants
mance risk and time risk appear to be more context specific. were recruited from the US only, and the results may not hold for
Overall, our study contributes to the previous literature by consumers from other cultural backgrounds. As individuals from
clarifying the effect of positive and negative factors on smart different cultural or professional backgrounds may have differing
home device adoption, as well as showing how the effects of perceptions of the risks associated with smart hone devices, future
these factors differ from previous literature in other contexts. studies are needed to identify and clarify these potential
Although not all hypotheses were supported, the use of a NVM differences.
proved valuable as about half of the variance for each dependent Notwithstanding this limitation, our study opens up many
variable was explained by our model (72.7% of perceived benefits, interesting opportunities for future studies. For example, future
45.1% of perceived risk, and 53.5% of intention to use smart home studies should examine how other risk factors influence indivi-
devices), providing strong support for our model. Also, our results duals’ overall perceived risk, as the dimensions identified in this
provide important insights regarding the effect of positive and study explain less than half of the variance. Further, studies could
negative factors on smart home device adoption. Here applying be conducted into how various factors moderate the relationship
NVM does not mean that every benefit/risk factor should be between perceived benefits/risk and adoption intention. Gewald,
significant. For example, when consumers perceive that financial et al.29 divided participants into different groups based on culture
risk is not an issue when adopting smart home devices, the insig- and found that the effects of perceived benefits/risk on BYOD
nificant effect of financial risk makes sense and is consistent with adoption intention were different across groups. It is possible that
consumers’ perceptions. Previous studies applying NVM have also individuals from different cultural backgrounds have different
found insignificant effects for certain benefit/risk factors. For perceptions of benefits and risk, which in turn may influence
example, Li, et al.28 found that only psychological risk has a their intention to adopt smart home devices. Studies should also
significant effect in the context of Italians’ health information be conducted to examine individuals’ post-adoption satisfaction.
seeking.
Conclusion
Implications for practice Individuals increasingly use smart home devices in their daily
lives.4 In order to examine how various benefit and risk factors
This study also has important practical implications. First, our influence individuals’ intention to adopt smart home devices, we
study shows that perceived benefits currently play the major role in developed a theoretical NVM that integrates both positive factors
the adoption of smart home devices, while perceived risk does not and risk factors. The model was tested with a US sample, and the
influence intention to use. Therefore, when conducting campaigns results provide partial support for our model.
to promote their devices, marketers should focus on benefits
provided by the devices. Our results can provide useful guidance
regarding which benefits to focus on in advertisements and pro- Funding
motions. For example, performance expectancy has a strong effect This work was supported by the Internal grant School of Engineering
on perceived benefit. Therefore, marketers need to emphasize how and IT, Murdoch University.
smart home devices can make individuals’ lives more convenient.
Further, as compatibility also has a significant effect, they should
emphasize that using (their) smart home devices will be compa- ORCID
tible with individuals’ lifestyles. Xuequn Wang http://orcid.org/0000-0002-1557-8265
JOURNAL OF COMPUTER INFORMATION SYSTEMS 445
References
22. Hsu C-L, Lin J-C-C. Exploring factors affecting the adoption of
1. Gartner. Gartner says 8.4 billion connected “things” will be in use Internet of Things services. J Comput Inf Syst. 2018;58(1):49–57.
in 2017, up 31 percent from 2016. Egham (UK): Gartner; 2017. doi:10.1080/08874417.2016.1186524.
2. Statista. Internet of Things (IoT) connected devices installed base 23. Coskun A, Kaner G, Bostan İ. Is smart home a necessity or a
worldwide from 2015 to 2025 (in billions); 2018 https://www. fantasy for the mainstream user? A study on users’ expectations of
statista.com/statistics/471264/iot-number-of-connected-devices- smart household appliances. Int J Des. 2018;12:7–20.
worldwide/. 24. Yang H, Lee W, Lee H. IoT smart home adoption: the importance
3. Statista. The Internet of Things (IoT) units installed base by of proper level automation. J Sensors 2018;2018:Article ID
category from 2014 to 2020 (in billions); 2018 https://www.sta 6464036.
tista.com/statistics/370350/internet-of-things-installed-base-by- 25. Fishbein M. Readings in attitude theory and measurement. New
category/. York (NY): Wiley; 1967.
4. McKinsey Global Institute. The internet of things: Mapping the 26. Lewin K, Dembo T, Festinger L, Sears PS. Level of aspiration. In:
value beyond the hype. McKinsey & Company; 2015. Hunt JM, editor. Personality and the behavior disorders. Oxford
5. Fernandes E, Jung J, Prakash A Security analysis of emerging (UK): Ronald Press; 1944. p. 333–378.
smart home applications. Proceedings of the 2016 IEEE 27. Li Y, Wang X. Online social networking sites continuance inten-
Symposium on Security and Privacy (SP); San Jose, CA; 2016. p tion: a model comparison approach. J Comput Inf Syst. 2017;57
636–54. (2):160–68. doi:10.1080/08874417.2016.1183448.
6. Barcena MB, Wueest C Security Response: insecurity in the 28. Li Y, Wang X, Lin X, Hajli M. Seeking and sharing health
Internet of Things; 2015 www.symantec.com/content/dam/syman information on social media: a net valence model and cross-
tec/docs/security-center/white-papers/insecurity-in-the-internet- cultural comparison. Technol Forecasting Soc Change.
of-things-15-en.pdf. 2018;126:28–40. doi:10.1016/j.techfore.2016.07.021.
7. Bastos D, Shackleton M, El-Moussa F Internet of Things: a survey 29. Gewald H, Wang X, Weeger A, Raisinghani MS, Grant G, Sanchez
of technologies and security risks in smart home and city envir- O, Pittayachawan S. Millennials’ attitudes toward IT consumer-
onments. Proceedings of Living in the Internet of Things: ization in the workplace. Commun ACM. 2017;60(10):62–69.
Cybersecurity of the IoT - 2018: IEEE; London (UK); 2018. doi:10.1145/3144574.
8. Lee MC. Factors influencing the adoption of internet banking: an 30. Kim D, Olfman L. Determinants of corporate web services adop-
integration of TAM and TPB with perceived risk and perceived tion: a survey of companies in Korea. Commun Assoc Inf Syst.
benefit. Electron Commer Res Appl. 2009;8(3):130–41. 2011;29:1–24.
doi:10.1016/j.elerap.2008.11.006. 31. Moore GC, Benbasat I. Development of an instrument to measure
9. Featherman MS, Pavlou PA. Predicting e-services adoption: a the perceptions of adopting an information technology innova-
perceived risk facets perspective. Int J Hum Comput Stud. tion. Inf Syst Res. 1991;2(3):192–222. doi:10.1287/isre.2.3.192.
2003;59(4):451–74. doi:10.1016/S1071-5819(03)00111-3. 32. Stone RN, Gronhaug K. Perceived risk: further considerations for
10. Wilson C, Hargreaves T, Hauxwell-Baldwin R. Benefits and risks the marketing discipline. Eur J Mark. 1993;27(3):39–50.
of smart home technologies. Energy Policy. 2017;103:72–83. doi:10.1108/03090569310026637.
doi:10.1016/j.enpol.2016.12.047. 33. Bauer RA. Consumer behavior as risk taking. In: Cox DF, editor.
11. GfK. Smart home: A global perspective. CES 2016. Las Vegas Risk taking and information handling in consumer behavior.
(USA): GfK; 2016. Cambridge (MA): Harvard University Press; 1967. p. 23–33.
12. Davis FD. Perceived usefulness, perceived ease of use, and user 34. Cunningham SM. The major dimensions of perceived risk. In: Cox
acceptance of information technology. MIS Quarterly. 1989;13 DF, editor. Risk taking and information handling in consumer beha-
(3):319–40. doi:10.2307/249008. vior. Cambridge (MA): Harvard University Press; 1967. p. 82–108.
13. Dong X, Chang Y, Wang Y, Yan J. Understanding usage of 35. Jacoby J, Kaplan LB. The components of perceived risk. In:
Internet of Things (IOT) systems in China: cognitive experience Venkatesan M, editor. Proceedings of the third annual conference
and affect experience as moderator. Inf Technol People. 2017;30 of the association for consumer research. Chicago (IL):
(1):117–38. doi:10.1108/ITP-11-2015-0272. Association for Consumer Research; 1972. p. 382–93.
14. Gao L, Bai X. A unified perspective on the factors influencing con- 36. Liew CS, Ang JM, Goh YT, Koh WK, Tan SY, Teh RY. Factors
sumer acceptance of Internet of things technology. Asia Pac J Market influencing consumer acceptance of Internet of Things technol-
Logist. 2014;26(2):211–31. doi:10.1108/APJML-06-2013-0061. ogy. In: Association IRM, editor. The internet of things:
15. Shin J, Park Y, Lee D. Who will be smart home users? An analysis Breakthroughs in research and practice. Hershey (PA): IGI
of adoption and diffusion of smart homes. Technol Forecast Soc Global; 2017. p. 71–86.
Change. 2018;134:246–53. doi:10.1016/j.techfore.2018.06.029. 37. Park E, Kim S, Kim Y, Kwon SJ. Smart home services as the
16. Kim Y, Park Y, Choi J. A study on the adoption of IoT smart next mainstream of the ICT industry: determinants of the adop-
home service: using value-based adoption model. Total Qual
tion of smart home services. Universal Access Inf Soc. 2018;17
Manag Business Excell. 2017;28(9–10):1149–65. doi:10.1080/ (1):175–90. doi:10.1007/s10209-017-0533-0.
14783363.2017.1310708. 38. Barbarossa C, Beckmann SC, De Pelsmacker P, Moons I,
17. Venkatesh V, Morris MG, Davis GB, Davis FD. User acceptance Gwozdz W. A self-identity based model of electric car adoption
of information technology: towards a unified view. MIS Quarterly. intention: a cross-cultural comparative study. J Environ
2003;27(3):425–78. doi:10.2307/30036540.
Psychol. 2015;42:149–60. doi:10.1016/j.jenvp.2015.04.001.
18. Kim HW, Chan HC, Gupta S. Value-based adoption of mobile 39. Yeh C-H, Wang Y-S, Yieh K. Predicting smartphone brand loy-
internet: an empirical investigation. Decis Support Syst. 2007;43 alty: consumer value and consumer-brand identification perspec-
(1):111–26. doi:10.1016/j.dss.2005.05.009. tives. Int J Inf Manage. 2016;36:245–57. doi:10.1016/j.
19. Yang H, Lee H, Zo H. User acceptance of smart home services: an ijinfomgt.2015.11.013.
extension of the theory of planned behavior. Ind Manag Data Syst. 40. Takayama L, Pantofaru C, Robson D, Soto B, Barry M Making
2017;117(1):68–89. doi:10.1108/IMDS-01-2016-0017. technology homey: finding sources of satisfaction and meaning in
20. Ajzen I. From intentions to actions: A theory of planned behavior. home automation. Proceedings of the 10th Conference on
Action control. Berlin, Germany: Springer; 1985. p. 11–39. Ubiquitous Computing; Pittsburgh, (PA) : ACM; 2012. p 511–20.
21. Hsu C-L, Lin J-C-C. An empirical examination of consumer 41. Aldás-Manzano J, Lassala-Navarré C, Ruiz-Mafé C, Sanz-Blas S.
adoption of Internet of Things services: network externalities The role of consumer innovativeness and perceived risk in online
and concern for information privacy perspectives. Comput banking usage. Int J Bank Marketing. 2009;27(1):53–75.
Human Behav. 2016;62:516–27. doi:10.1016/j.chb.2016.04.023. doi:10.1108/02652320910928245.
446 X. WANG ET AL.
42. Rakhi T, Mala S. Adoption readiness, personal innovativeness, 47. Lindell MK, Whitney DJ. Accounting for common method var-
perceived risk and usage intention across customer groups for iance in cross-sectional research designs. J Appl Psychol.
mobile payment services in India. Internet Res. 2014;24(3):369– 2001;86:114–21.
92. doi:10.1108/IntR-12-2012-0244. 48. Ringle CM, Wende S, Will S. SmartPLS 2.0 (M3) Beta. Hamburg;
43. Müller-Seitz G, Dautzenberg K, Creusen U, Stromereder C, 2005.
Müller-Seitz G, Dautzenberg K, Creusen U, Stromereder C. 49. Hulland J. Use of partial least squares (PLS) in strategic
Customer acceptance of RFID technology: evidence from the management research: a review of four recent studies.
German electronic retail sector. J Retailing Consumer Serv. Strategic Manag J. 1999;20(2):195–204. doi:10.1002/(ISSN)
2009;16(1):31–39. doi:10.1016/j.jretconser.2008.08.002. 1097-0266.
44. Hargreaves T, Wilson C, Hauxwell-Baldwin R. Learning to live in 50. Henseler J, Ringle CM, Sarstedt M. A new criterion for assessing
a smart home. Building Res Inf. 2017;46(1):127–39. discriminant validity in variance-based structural equation mod-
45. Fishbein M, Ajzen I. Predicting and changing behavior: The rea- eling. J Acad Marketing Sci. 2015;43(1):115–35. doi:10.1007/
soned action approach. New York (NY): Psychology Press; 2010. s11747-014-0403-8.
46. Podsakoff PM, MacKenzie SB, Lee J-Y, Podsakoff NP. 51. Tirado Herrero S, Nicholls L, Strengers Y. Smart home technol-
Common method biases in behavioral research: a critical ogies in everyday life: do they address key energy challenges in
review of the literature and recommended remedies. J Appl households? Curr Opin Environ Sustainability. 2018;31:65–70.
Psychol. 2003;88(5):879–903. doi:10.1037/0021-9010.88.5.879. doi:10.1016/j.cosust.2017.12.001.
JOURNAL OF COMPUTER INFORMATION SYSTEMS 447
APPENDIX: MEASUREMENT
Intention to use smart home devices (developed for this study using the guidelines from Fishbein and Ajzen’s reasoned action approach45)
Int1 I would like to use smart home devices.
Int2 I expect to use smart home devices.
Int3 I hope to use smart home devices.
Int4 I can see myself using smart home devices.
Perceived benefits (adapted from Li, et al.28)
PB1 Using smart home devices offers me a lot of advantages.
PB2 Using smart home devices is beneficial.
Performance expectancy (adapted from Venkatesh, et al.17 and Hsu and Lin21)
PE1 Smart home devices are very useful.
PE2 Smart home devices improve the quality of household management.
PE3 Using smart home devices would enable me to accomplish tasks more quickly.
Effort expectancy
EE1 Smart home devices are easy to use (from Venkatesh, et al.17)
EE2 Smart home devices are easy to find in the marketplace (developed for this study).
EE3 It will be quick for me to learn how to use smart home devices (developed for this study).
EE4 It will be easy for me to learn how to use smart home devices (from Venkatesh, et al.17)
EE5 Operation of smart home devices is clear and understandable (from Venkatesh, et al.17)
Compatibility (adapted from Moore and Benbasat31)
COM1 Using smart home devices fits with my approach to managing things at home.
COM2 Using smart home devices fits well with the way I like to live.
COM3 Using smart home devices fits into my lifestyle.
Image (adapted from Moore and Benbasat31)
Image1 People who use smart home devices have a more prestigious image than people who do not.
Image2 People who use smart home devices have a high profile.
Image3 Use of a smart home device presents a positive image to other people.
Image4 Having a smart home device would be a status symbol.
Perceived risk (adapted from Li, et al.28)
PR1 I expect that using smart home devices would present me with problems that I just don’t need.
PR2 The benefits of using smart home devices are unlikely to compensate for the cost, time and effort of using them.
Privacy risk (from Featherman and Pavlou9 and Li, et al.28)
PriR1 If I use a smart home device, I would lose control over the privacy of my personal data.
PriR2 My personal information will be less confidential if I use a smart home device.
Security risk (from Featherman and Pavlou9)
SR1 The security systems built into smart home devices are not strong enough to protect my information.
SR2 Internet hackers (criminals) might take control of my information if I use a smart home device.
Performance risk (adapted from Featherman and Pavlou9)
PerfR1 Smart home devices might not work well.
PerfR2 Smart home devices might create problems in my home.
PerfR3 Smart home devices might not perform as well as promised.
PerfR4 Smart home devices might not meet my requirements.
Time risk (developed for this study based on Featherman and Pavlou9)
TR1 A lot of my time could be wasted in setting up smart home devices.
TR2 I could lose a lot of time learning how to use smart home devices.
TR3 I would need to invest a lot of time to get smart home devices to work well in my home.
Financial risk (developed specifically for this study)
FR1 Smart home devices are expensive to buy.
FR2 Setting up smart home devices costs a lot of money.
FR3 Keeping smart home devices running is a financial burden.