User Expectations for Comfort,Discomfort,and Wearability of Wrist-Worn Devices

Abstract

Comfort has been extensively studied and identified as one of the most important factors in the adoption of wearable devices, yet there is still no single commonly accepted definition. Researchers have also examined connections between comfort and discomfort, but less work has connected to the concept of wearability. Extending a 1996 study from Zhang, Helander, and Drury, we conducted three between-subjects surveys designed to account for different definitions, better distinguish between the different constructs, and detect device-specific differences (watch, fitness tracker, smart watch). We found that comfort and wearability may be interchangeable in wrist-wearable research, but discomfort should be considered separately. Device function impacted user expectations for comfort, discomfort, and wearability, meaning that results from research conducted on watches may differ from that of smart watches.

Introduction

The development of wearable technologies is growing rapidly, with contemporary products featuring a combination of sensors and computing devices embedded in clothing and fashion accessories, such as activity monitoring bracelets or smart watches (Nieroda et al., 2018). Wearable devices have a range of potential benefits including health, fitness tracking, notification management, and more. These devices are becoming increasingly popular, with the global wearables market expected to reach $150 billion by 2026 (Ometov et al., 2021). Despite the advantages and growth potential of wearable devices, their adoption has been slower than expected. Previous research identified comfort as one of the most important factors in the adoption of wearable devices (Kalantari, 2017). Thus, understanding factors affecting comfort is a prerequisite for their widespread adoption (Gemperle et al., 1998), yet there is currently no single commonly accepted definition of comfort in the context of wearable devices.

Comfort and discomfort have been interchangeably used in a range of clinical settings and wearable evaluations (Pearson, 2009), but research indicates that comfort and discomfort are both multidimensional constructs affected by distinctly different variables (de Looze et al., 2003; Knight & Baber, 2005; Zhang et al., 1996). Less work has connected to the concept of wearability. Although there is no single commonly accepted definition, several studies characterized wearability as the factors that affect the degree of comfort the wearer experiences while wearing a device, including physical, psychological, and social aspects (Dunne et al., 2014; Dunne & Smyth, 2007; Gemperle et al., 1998). All of these constructs (comfort, discomfort, and wearability) are complex and multidimensional.

The role of device type and associated functions cannot be ignored. There is a substantial body of research examining the comfort of chairs and seats (e.g., Kyung et al., 2008; Kyung & Nussbaum, 2008), built largely from Zhang et al.’s (1996) seminal work and expanding beyond the office into other domains (e.g., automotive seating). However, the difference between user experiences of comfort and discomfort between sitting and wrist-worn devices is likely substantial.

More recent research for wearable devices resulted in tools, such as Knight and Baber’s (2005). Since we expect that device function influences users’ perceived important factors for comfort, discomfort, and wearability, we isolated device function as a variable, recognizing the range of devices on the market today that are already widely available and familiar to end users: watches, fitness trackers, and smart watches.

The goals of this work are to:

1. Comprehensively account for factors impacting comfort, discomfort, and wearability in wrist-worn devices.

2. Understand the relationship between comfort, discomfort, and wearability for wrist-worn devices.

3. Identify differences between user expectations for different types of wrist-worn devices: watches, fitness trackers, and smart watches.

Our approach aims to provide recommendations for evaluating comfort, discomfort, and wearability of various types of wrist-worn devices. The results can be applied to both practical use cases and theoretical knowledge.

Methods

Replicating Zhang et al. (1996), we conducted a series of three IRB-approved between-subjects surveys: a questionnaire study, a validation study, and a classification study. In the first two studies, participants were screened to ensure they regularly wore devices on their wrist, so we could capture the expectations of users familiar with each device type. This is in line with Knight and Baber’s (2005) recommendation that assessments “should involve participants who use such devices or for whom the device is being developed. . .[for] a more meaningful and useful assessment of the comfort of the device.”

In the Questionnaire and Validation Studies, participants were grouped into subgroups based on the wrist-worn device they reported wearing most often:

• Watch: a wrist-worn device that tracks time

• Fitness tracker: a wrist-worn device that records a person’s daily physical activity, together with other data relating to their fitness or health, such as the number of calories burned, heart rate, etc.

• Smart watch: a wrist-worn device with a touchscreen display that integrates with other devices and enables functions like fitness tracking, text messaging, music control, and more

Each participant was also randomly assigned one of three descriptors: comfort, discomfort, or wearability. This resulted in nine sub-groups, one for each combination of device (watch, fitness tracker, smart watch) and descriptor (comfort, discomfort, wearability). We made three significant adaptations to the original Zhang et al. (1996) study to address our goals: (1) we included a third construct of “wearability” (in addition to comfort and discomfort), (2) we further subdivided participants into groups based on the specific type of wrist-worn device that they wore most often, (3) we extended the list of words to include additional terms that might be relevant for wearable computers.

Questionnaire Study

The questionnaire study (n = 1,240) aimed to generate a comprehensive list of descriptors for comfort, discomfort, and wearability of wrist-worn devices. We followed the methods described in Zhang et al. (1996) closely, resulting in three survey sections. First, we conducted a general assessment of factors important in wrist-worn device design, with an open-ended question asking participants to rank order their opinions and impressions. Second, participants generated descriptors associated with their assigned descriptor (comfort, discomfort, wearability) for their most-used wrist-worn device type, with an open-ended question. Third, participants evaluated experimenter-generated descriptors by rating the relatedness of 176 words seeded from previous research (Knight & Baber, 2005; Zhang et al., 1996) to the assigned descriptor for their device type. Relatedness was rated as “related (positively or negatively),” “not related,” or “not sure.”

Validation Study

In the validation study (n = 1,214), participants rated the 36 most related words from the questionnaire study (those rated as “related” by at least two-thirds of participants) in terms of how closely they related to the assigned descriptor for the participant’s subgroup’s device type. Closeness was rated on a four-point scale from “very closely related” to “not related at all,” with an additional “don’t know” option. Only words rated as “very closely related” or “closely related” by at least 70% of the participants in a subgroup were retained for the classification study.

Classification Study

In the classification study (n = 28), participants familiar with rating scales completed pairwise comparisons of all possible pairs of the 36 descriptors retained from the validation study. For each pair, participants rated the similarity of the words on a scale from 1 (totally different) to 7 (almost the same). These comparisons were used to generate a similarity matrix for further analysis.

Findings

Questionnaire Study Results

The open-ended questions in Sections 1 and 2 of the questionnaire study aimed to identify any additional descriptors missing from our original list of 176 words. Approximately half of the 4,355 unique descriptors generated by participants were accounted for in the original list (2,184 responses, 50.2%).

Another 32.7% referenced unspecified device functionality (e.g., “it works,” 316 responses), specific applications (824 responses), or design characteristics (e.g., waterproofing, attachments, changeable bands, specific materials; 284 responses). As the goal was to identify generalizable features across wrist-worn devices, overly specific details like particular materials were avoided to minimize the influence of trends that may change over time. Responses relating to device marketing (109 responses) or age (30 responses) were also excluded. Using this exclusion criteria, all valid responses within the scope of the study were already captured by our existing descriptor list.

In Section 3, participants rated the relatedness of the 176 experimenter-generated descriptors to the assigned construct (comfort, discomfort, wearability) for their most-used device type. To determine which words would move forward to the validation study, we used the criterion from Zhang et al. (1996) that more than two-thirds of participants had to consider the word “related.” This process resulted in a list of 36 words. Notably, none of the 36 words that met the criterion for the validation study had inherently negative connotations.

Validation Study Results

The validation study provided more granularity on the 36 words selected from the questionnaire study. Participants rated how closely each word related to the assigned descriptor (comfort, discomfort, wearability) for their most-used device type. To be retained for the final classification study, at least 70% of participants in a subgroup had to rate the word as “very closely related” or “closely related.”

Classification Study Results

In the classification study, participants rated the similarity of all possible pairs of the 36 words retained from the validation study. These ratings were used to generate similarity matrices, which were then analyzed using factor analysis. Factor analyses were conducted for the three device types (watch, fitness tracker, smart watch) and the three descriptors (comfort, discomfort, wearability). For fitness trackers (30 words) and smart watches (34 words), a five-factor solution emerged: function/reliability, emotion, safety/security, adjustability, and appearance/fit. For watches (14 words), only three factors were identified: function, trustworthiness, and appearance. When examining the descriptors across device types, comfort (21 words) and wearability (26 words) yielded the same five factors: function, safety/security, adjustability, satisfaction, and appearance/fit. Discomfort (15 words) was associated with just three factors: function, appearance/fit, and security/reliability.

Discussion

The findings from this series of studies offer several insights into how people perceive and evaluate wrist-worn devices. Comfort and wearability were closely related based on the factor analysis results, with both constructs associated with the same five factors (function, safety/security, adjustability, satisfaction, appearance/fit). Discomfort, on the other hand, was distinct and only associated with three factors (function, appearance/fit, security/reliability). This suggests that while comfort and wearability may be interchangeable concepts in the context of wrist- wearable research, discomfort should be treated as a separate construct and measured independently.

Across all device types and constructs, function consistently emerged as a key factor. Even though comfort research often focuses heavily on the physical experience of wearing a device, our results highlight that the perceived capabilities and functions of wrist- worn devices also significantly impact users’ comfort perceptions. Fitness trackers and smart watches yielded very similar five-factor structures, while basic watches only produced three factors and seemed to stand apart. This supports and extends the findings of Knight and Baber (2005), which emphasized the role of psychological factors in addition to physical ones. An important implication is that researchers should consider and account for the expected functions of devices when designing studies on the comfort, discomfort, or wearability of wrist-wearables, as this appears to be a universally important factor shaping user perceptions and experiences. It is critical that researchers identify the expected function of devices if known, or account for the potential for function to sway user expectations. For research purposes, smart watches and fitness trackers may be treated similarly, but watches should be considered a distinct category.

Perhaps the most striking finding from this research was the absence of negatively valenced terms in the descriptors participants associated with all three constructs (comfort, discomfort, wearability) across the three device types (watches, fitness trackers, smart watches). This contrast previous research on seating comfort, such as the work of Zhang et al. (1996) where discomfort emerged as a major factor alongside comfort, or Knight and Baber’s (2005) study on wearable computers which included negative factors like harm and anxiety. A possible explanation for this difference is that wrist-worn devices, unlike chairs, are generally seen as value-added accessories. If the discomfort or inconvenience of wearing one outweighs the benefits, users can easily remove the device. Chairs, in contrast, are typically a necessity, and even an uncomfortable chair may be preferable to not having one at all. People may choose to wear wrist devices for the value they provide, but this value can often be attained through alternative means. If a device fails to deliver a comfortable user experience, discarding or replacing it may be seen as an easier solution than tolerating discomfort.

The absence of negative descriptors resulted in more granular, differentiated factor structures in our research compared to previous studies. Without broad negative factors like discomfort or anxiety, the other factors could be expressed in finer detail. This finding highlights the importance of considering the specific type of device or technology being studied, as the fundamental nature of the user’s relationship with it can greatly impact how they perceive and evaluate the experience of using it.

Conclusions

We aimed to categorize factors impacting comfort, discomfort, and wearability in wrist-worn device design, understand the relationship between these constructs, and identify differences in user expectations across device types. Our results suggest that comfort and wearability are closely related for wrist-worn devices and may be interchangeable in research, while discomfort is a distinct construct that should be measured separately.

Device type and associated functions impact user perceptions and expectations. Smart watches and fitness trackers yielded similar factor structures, while basic watches stood alone, implying that research on wrist-wearables may not need to distinguish between smart watches and fitness trackers but should consider watches separately. Across all device types, function emerged as a key factor, highlighting the importance of accounting for device capabilities when studying the user experience of wrist-worn technologies.

Interestingly, participants only associated the wrist-worn devices with positive or neutral terms, contrasting with prior research on office chairs that included negative descriptors. This suggests wrist-wearables are viewed as value-added devices, and if that value is eclipsed by discomfort or other negative factors, users may simply abandon them. The positive framing of descriptors related to wrist-worn devices resulted in a more granular factor structure compared to previous studies.

Our findings provide a foundation for evaluating the comfort, discomfort, and wearability of various types of wrist-worn devices. By considering the multidimensional nature of these constructs and the impact of device type and function, researchers and designers can more comprehensively assess and improve the user experience of wrist-wearable technologies.