Keywords

1 Introduction

Wearable technology presents an interesting design problem in that wearables are not clearly defined as either “fashion accessory” or “technological innovation.” This ambiguity makes it difficult to predict which factors determine whether a wearable, like FitBit activity trackers or Spectacles by Snap, will be adopted by a broad audience of potential wearers. For her dissertation, author Kelly [8] systematically created and validated the WEAR Scale, a 14-item questionnaire for predicting the social acceptance of a wearable device. As she followed steps for validation outlined by DeVellis [4], she gathered data about wearable acceptance from four studies. While she successfully validated the WEAR Scale, her data from the four studies offered varied results regarding the extent to which fashion and aesthetics affect the adoption of wearable devices. The current paper explores her results with a focus on that particular issue, looking more closely at the aesthetic considerations that have been identified as important in literature, as well as at data by experts on fashion and in the area of wearables, themselves.

Up to this point in the history of wearables, the focus of innovation has been in functionality despite the growing calls for fashion and aesthetic considerations for wearable tech [6, 11, 14, 16]. This push for more fashionable wearable tech has produced worry for some, however. Edwards [3] expressed a fear of the marketing drive for fashion audiences limiting the innovation of the technology. His concern seems to have been realized in the success and failure stories of the wearable tech that have surfaced since Edwards’s writing in 2003. Google Glass was briefly released for public beta testing from 2011 to early 2013 [11] and the Glass Explorer program was officially shut down in January of 2015 [2]. As highlighted by Page [11], the public welcomed Glass with incredible enthusiasm, but ultimately the wearable receded from the public eye as a highly unfavorable device. Google’s failure with Glass was a multifaceted phenomenon. “Glassholes,” as the early Glass adopters came to be called, eventually found little functionality in the device, while some others reported the failure to be largely due to the unattractiveness of Glass[11]. Still others noted that Glass was actually quite attractive in its design, insisting that developers simply needed a lesson in how to move from “attractive” to “fashionable” [16]. Additionally, the computer-enhanced eyewear raised a number of privacy concerns both for potential users in the medical field [17] and for anyone who had to wonder whether they were being recorded without warning [2].

However, the success of other devices, such as the FitBit and JawboneUP activity tracker bracelets and the Apple Watch, is proof that wearables do interest the public. In order to make the full shift from the attention of early adopters and technology-enthusiasts to the general public, however, wearable developers must gain a better understanding of how to market their devices. In her previous work, Kelly [8] defined wearables as computers or electronic devices displayed voluntarily in a manner similar to fashion accessories, like jewelry or sunglasses. For such devices, she noted, social acceptance is imperative because the indulgence in wearable technology is a very public affair.

As noted above, there is a lot of evidence pointing toward the importance of a fashionable, aesthetically pleasing wearable device. As noted by Aspers and Godart [1], fashion persists throughout contemporary life. It is of little surprise, therefore, that wearable devices are also influenced, and often molded by fashion. Besides this more philosophical reflection of the influence of fashion, it is also notable that like fashion accessories which are also worn publically, wearables are a form of nonverbal communication. Wearables establish an identity for those who don them, are expected to communicate positive messages, and display membership to likeable social groups[1, 7]. Fashion items must be consistent with the person’s self-image to be acceptable on a personal level[7], and can even elicit dominance and subordinance behavior interaction in dyads, if the dress of one member signifies membership to a higher social class[9].

When author Kelly conducted research on wearables to construct the WEAR Scale[8], she gathered data from multiple sources:

  1. 1.

    Interviews with potential users of wearables (N = 9)

  2. 2.

    Survey of experts in the field of wearables (N = 3)

  3. 3.

    Survey 1 of potential users of wearables (Study 1, N = 221)

  4. 4.

    Survey 2 of potential users of wearables (Study 2, N = 306)

To explore the issue of whether fashion aesthetics contributes to a decision to adopt a wearable, it is worth examining findings related to aesthetics in each of these. At each step of the data gathering process, the focus was on determining which items (questions) for the WEAR Scale would be most predictive of social acceptability of a wearable device. While Kelly [8] focused on the construct of social acceptability of a wearable, a factor which she noted would be necessary for broad adoption, is it worth examining the data further to see whether “fashionable” might be an additional construct that is required for adoption.

2 Interviews and Experts

In initial interviews with the WEAR Scale’s target population, people aged 18–30, fashion was indeed a consideration for some when assessing the social acceptability of a wearable. Fashion was mentioned as one purpose that a wearable could take on as well as a somewhat-important criterion for choosing a wearable. Interestingly, fashion considerations were not a popular answer among interviewees, as they believed that a “fashionable” wearable would wane in popularity when it is no longer trendy and that fashion isn’t important to all the potential users of wearable devices. Thus, overall, despite the trends described above, the interview participants de-emphasized the role of fashion in wearable adoption. This was expected to be an artifact of the small sample size (N = 9), initially, but as discussed later in this paper, it is possible that this trend first noted in these interviews is an idea shared by many.

Kelly [8] also vetted the WEAR Scale items with three experts in the field of wearable technology for review. Two scale items, This device is stylish and This device is fashionable, related directly to fashion and were rated as being very relevant to social acceptability of a wearable. The experts rated the question category that these items belonged to, Aesthetics, as being quite relevant overall, averaging 1.60 out of 3, where 1 is very relevant, 2 is somewhat relevant, and 3 is not relevant to social acceptability of a wearable. This rating was among the best, with only the category Others’ thoughts being rated as more relevant. Additionally, some of the experts offered comments which emphasized a focus on the aesthetics of the device.

It is worth exploring the trend uncovered in the population interviews further; does fashion matter when talking about the social acceptability of a device? Yang et al. [18] found that both social image and visual attractiveness are important factors in the public perception of wearables. If that is true, they may be part of the same construct, social acceptability. If not, the question of whether fashion or social acceptability is more important for wearable adoption is worth examining. By looking further at the procedural construction of the WEAR Scale, specifically in the journey of the question items which pertain to aesthetics, the authors seek to answer such questions.

3 Study 1

The WEAR Scale initially consisted of 50 question items and was established using the results of a thorough literature review, initial population interviews, and expert item review, all three of which were briefly described previously [see 8]. Not discussed in her previous work are the conclusions about fashion’s role in the social acceptability of a wearable which can be pulled from her data. To test the validity of the WEAR Scale, the researcher recruited 221 participants from the Iowa State University student body to take a survey which asked those original 50 WEAR items, as well as demographic questions and questions which captured information on related concepts, such as Usability and Ease of Use and technology adoption preferences. Validation of the WEAR Scale also had implications for the relationships between concepts thought to drive the social acceptability of wearable devices.

3.1 Methods

Participants.

All participants were required to be aged 18 to 30, because this is the population for which the WEAR Scale was developed. This population was targeted because young adults are generally considered the largest age demographic for technology adoption[15]. Of the 221 participants whose data was used in analysis, 57.5% were male (N = 127), 42.1% were female (N = 93), and 1 did not answer. The majority of the sample had some college (N = 150, 67.9%), 17 had graduated from high school (7.7%), 18 had graduated from college (8.1%), 21 had some graduate work (9.5%), and 15 held Masters, Ph.D., or other advanced degrees (6.8%).92.3% of participants did not identify as Hispanic or Latino/a (N = 204, 1 unanswered), 82.4% of participants identified as White (N = 182), 10.9% identified as Asian (N = 24), 3.2% identified as Black (N = 7), less than 3% identified as Native Hawaiian/Pacific Islander or American Indian/Alaska Native (2.3%, N = 5), and 3 participants did not supply their race.

Tasks.

After reviewing the informed consent approved by the Institutional Review Board, participants completed the survey online, consisting of the 50 WEAR items as well as demographic and validation items. Participants were presented with two photos of the LG Tone + HBS-730 Wireless Bluetooth Stereo Headset Neckband (Fig. 1), and a description of the functions and purpose of the wearable. The Bluetooth headset was chosen for this study because it was expected to evoke useful variability – the neckband did not resemble a common accessory, but it was an existing product that participants might own themselves, or may have seen on other people.

Fig. 1.
figure 1

The pictures shown to participants along with a description of the functions and purpose of the device. Left: vapingunderground.com/threads/lg-tone-hbs-730-wireless-bluetooth-stereo-headset-neckband-style-hbs730.73983/. Right: www.youtube.com/watch?v=ucFnZS8sOww

Full size image

Dependent Variables.

While many different dependent variables, including device WEAR Score and participant optimism, were collected, those of interest in this paper are those WEAR items relating to aesthetic (Table 1). These were originally included because the social acceptability of wearable devices was assumed to be related to theories of dress and fashion, as revealed in the literature review and the expert review. These variables also were used in the second user study, detailed below.

Table 1. The initial WEAR scale’s aesthetic items loading values and the component each loaded on, when applicable, based off data collected using the Bluetooth Headset. The first three items loaded significantly onto a single component, while the remaining four did not.
Full size table

3.2 Initial Results

Of the 302 participants, 221 gave usable data. Of the 50 WEAR items, 31 items were found to significantly load onto one of two components. In this phase of testing, many of the aesthetics items (Table 1) were found to contribute to the understanding of the “social acceptability of a wearable device.”

Two components or factors were identified, and items were attributed to a factor if orthogonal (Varimax) rotation resulted in a loading of at least 0.30 on one, but not more than one, factor. When an item loads on more than one factor, it becomes the responsibility of the researcher to assign said item to a factor – a process which adds subjectivity to the structure of the survey. Because a simple structure was desired for WEAR, an item loading on more than one factor was considered complex and was excluded. Before discussing the implications of this result for aesthetics, it is useful to consider Study 2.

4 Study 2

Because the WEAR Scale is intended to be valid for testing the social acceptability of any wearable device, the preceding Bluetooth study needed to be repeated with additional devices, with the intention of arriving at a common solution for all devices tested. As in the Bluetooth study, wearables were sought that would evoke sufficient variability. Therefore, Apple Watch (Fig. 2) and Google Glass (Fig. 3) were selected, with the Watch likely to be considered more acceptable due to its similarity to the common wrist watch, and Glass less acceptable because it is a head-worn device that initially raised some controversy.

Fig. 2.
figure 2

The image of Apple Watch, shown to participants during the second user evaluation of the WEAR Scale. Left image: www.apple.com/shop/buy-watch/apple-watch-sport. Top right image: www.macworld.co.uk/how-to/apple/guide-phone-calls-on-apple-watch-3607555/. Bottom left image: www.theverge.com/2014/9/9/6124253/apple-watch-hands-on-video-photos.

Full size image
Fig. 3.
figure 3

The image of Google Glass, shown to participants during the second user evaluation of the WEAR Scale. Left image: www.pcadvisor.co.uk/feature/gadget/google-glass-release-date-uk-price-specs-3436249/. Middle and right image: fortune.com/2015/12/28/new-google-glass/.

Full size image

4.1 Methods

Participants. As with the first study, all students at Iowa State University were invited via email to complete the survey online and were required to be aged 18 to 30 to participate. Of the 306 participants whose data was used in analysis, 56.5% were female (N = 173), 41.5% were male (N = 127), 1.6% didn’t identify as male or female (N = 5), and 1 participant didn’t answer. Again, the majority of the sample had completed some college (N = 234, 76.5%), followed by having completed some graduate work (N = 29, 9.5%), holding a high school diploma (N = 17, 5.6%), holding a college degree (N = 17, 5.6%), and holding Masters, Ph.D., or other advanced degree (N = 8, 2.6%). One participant did not respond to this question. 94.1% of participants did not identify as Hispanic or Latino/a (N = 288), 88.6% of participants identified as White (N = 271), 7.8% identified as Asian (N = 24), less than 3% identified as Black or American Indian/Alaska Native (1.3%, N = 7), none (N = 0) identified as Native Hawaiian/Pacific Islander, and 4 participants did not supply their race.

Tasks. A description of each device’s functions and purpose again accompanied the images of Apple Watch and Google Glass. Because of the length of the survey, participants answered two surveys one week apart: the questions with regard to Watch and the questions with regard to Glass (with the demographic and validation items split between the two administrations). To account for the possibility of order effects, half of the sample saw the Apple Watch survey and accompanying validation items first and the other half saw the Google Glass survey and accompanying validation and demographic items first.

4.2 Initial Results

Again, some data were dropped due to participant age, short survey time, or skipped WEAR items, leaving 306 instances of usable data. Four components or factors were identified and items were attributed to a factor if Oblim in rotation resulted in a loading of at least 0.30 on one, but not more than one, factor. Just as for the first study, researchers desired simple structure for WEAR, so an item loading on more than one factor was considered complex and was excluded.

Of the 50 WEAR items, 34 items were found to significantly load onto one of four components for Apple Watch, and 30 loaded significantly onto one of four components for Google Glass. By examining the number of items loading, alone, one can see that some items needed to be thrown out due to inconsistency. Additionally, by examining Table 2, which details the loadings of the aesthetics items for Apple Watch, and Table 3, which details the same for Google Glass, one can see that there is inconsistency between items across components. Prior to this phase of testing, many of the aesthetics items (Table 1) were found to contribute to the understanding of the “social acceptability of a wearable device.” However, the noted inconsistencies in how these items filter onto components speak to the importance of aesthetics in the social acceptability of a wearable device.

Table 2. The initial WEAR scale’s aesthetic items loading values and the component each loaded on, when applicable, based off data collected using Apple Watch. The first three items loaded significantly onto a single component, while the remaining four did not.
Full size table
Table 3. The initial WEAR scale’s aesthetic items loading values and the component each loaded on, when applicable, based off data collected using Google Glass. The first two items loaded significantly onto a single component, while the remaining four did not.
Full size table

5 Discussion

When comparing aesthetics items across surveys for Apple Watch (Table 2) and Google Glass (Table 3), it is apparent that none of the aesthetic items loaded consistently on a component, which means that any attempts to wrangle the items into a specific component or category is a very subjective choice. When deciding whether a question item belongs with all of the others that currently fit into a component, researchers add their own personal bias to the survey. This isn’t to say that surveys in which this method of organization has been used are not valid, but rather that author Kelly did not want to complicate her survey.

Additionally, when comparing between the surveys for all three wearable devices, it is clear that none of the aesthetics items loaded consistently on any one component. It is important that the items load onto the same items when tested in different situations so that a survey continues to measure a construct reliably in different contexts. Kelly, by comparing the results in the two studies described above, determined that a 14-item, two-factor solution was the most appropriate structure for the WEAR Scale. All items included in this solution loaded consistently across surveys, and average Cronbach’s alpha for the final solution was 0.84. [8 gives more detail about this process.]

Thus, the WEAR Scale was determined to be composed of the two factors Aspirational Desires and Social Fears. Despite the significance of aesthetics and fashion in the literature and in the initial analysis of the scale, this pattern of the importance of aesthetics in social acceptability of wearable devices was not stably supported in the WEAR Scale. The factor analysis suggested that aesthetics might be a construct completely separate from social acceptability, which, based off the responses of the three experts called upon to give insight to the WEAR scale, is not as clear for those developing wearable devices or researching their impact on and influence by human behavior. Reflecting on the interview data, while it was true that aesthetics did appear in respondents’ answers, the responses more often pointed for a need of an absence of aesthetics as a consideration in social acceptability. Indeed, it is telling that participants indicated their belief that the trendiness of a wearable would negatively correlate with the lifespan of the device.

Another important lesson from the WEAR Scale is spelled out in the second factor, Social Fears. Arguably the biggest reason for Google Glass’s downfall was the creepiness of the device [12]. Glass had video-capture functionality, and the only indication of whether the camera was active was a tiny light at the corner of the device’s frame. Because of the inconspicuousness of Glass’s recording function, it was quickly banned from bars due to privacy concerns, and from places like movie theaters and casinos where unauthorized recording is not condoned. Based on research demonstrating a preference for the socially acceptable variety of uniqueness [10], it could be argued that wearables are endeavoring to walk a very narrow path of uniqueness which falls at the edge of socially acceptable uniqueness, such as owning thrifted items, and socially unacceptable uniqueness, such as disregarding others’ reactions to a behavior or item.

An alternative explanation of the threat sometimes caused by wearables comes out in a discussion of how fashion can be used. By observing what someone is wearing, it is possible to discern social rank, as shown in work by Kraus and Mendes [9]. Additionally, their research showed that in pairs, when it is clear that one is of a lower social class than his or her partner, the person of lower social standing will experience increased threat vigilance and an alteration of hormone levels, which is to say that knowing someone holds social dominance over oneself causes distress. Regardless of whether the social fears are due to the surreptitious functions of some wearables, or the symbol of power that they instill, it is clear that if a wearable is to survive among consumers, it must sufficiently deal with such social considerations.

6 Conclusion

It was hypothesized that fashion would be a central concept to the social acceptability of a wearable; however, it is apparent that such aesthetic considerations are actually separate from, but related to, a wearable device’s social acceptance. Perhaps Wasik [16] had greater insight than he’d intended when he noted that wearables should either be hyper-attractive or microscopic. It would appear that subtlety is the most important aesthetic consideration when designing a socially acceptable wearable, as seen in the work done by Toney, Mulley, Thomas, and Piekarski [13] during the development of their e-SUIT.

Regarding the eventual exclusion of the aesthetic items that were originally part of the WEAR item pool, it is clear that there are discrepancies between experts in the field and the data obtained in empirical research. Aesthetics, though related to social acceptability, does not fall under the same construct. Further research should be done to determine exactly how these constructs are related.

Aesthetic appeal cannot outweigh the fears and concerns associated with wearables, which go beyond those noted with regards to Glass. These results suggest that it is important for industry designers to reduce the focus put onto aesthetics of such devices when designing for social acceptability, making sure to heed the potential social fears that such a device could trigger. If a wearable device has the potential to negatively impact the welfare of others or rejects social conventions, such as manners, it will lose favor with the public.