Barcelona, Spain | September 3, 2018 MobileHCI 2018 Workshop on Socio-Technical Aspects of Text Entry Put a Ring on It: Text Entry Performance on a Grip Ring Attached Smartphone Monwen Shen Abstract Gulnar Rakhmetulla This paper presents results of a study investing effects Ahmed Sabbir Arif of grip rings on text entry. Results revealed that grip Human-Computer Interaction Group rings do not affect text entry performance in terms of University of California, Merced speed, accuracy, or keystrokes per character. It then Merced, CA, USA reflects on future research directions based on the mshen6@ucmerced.edu results and observations from the study. The purpose grakhmetulla@ucmerced.edu of this work is to stress the necessity of classifying and asarif@ucmerced.edu evaluating low-cost mobile phone accessories. Author Keywords Bunker ring; grip ring; smartphone; accessories; gadgets; text entry; input; interaction. ACM Classification Keywords H.5.2 User Interfaces (D.2.2, H.1.2, I.3.6): Input devices and strategies (e.g., mouse, touchscreen); H.1.2. Models and Principles: User/Machine Systems. Introduction Low-cost mobile phone accessories, such as portable Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed power banks, panoramic pods, selfie sticks, different for profit or commercial advantage and that copies bear this notice and the full citation types of lenses for photography, screen magnifiers, VR on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s). headsets, cases, grip rings, PopSockets, etc., are Copyright held by the owner/author(s). becoming increasingly popular among mobile users [6], MobileHCI, 2018 Barcelona, Spain. likely due to their affordability. With mobile phones, these accessories are also becoming ubiquitous. Hence, it is essential that these devices are designed with due 6 Barcelona, Spain | September 3, 2018 MobileHCI 2018 Workshop on Socio-Technical Aspects of Text Entry consideration of human factors. Since not much work to four-loop attachment that enables thumb-free grip of has focused on this, we review and evaluate the mobile phones (Figure 4). Unlike most other grips, it effectiveness and usability of different mobile phone cannot be used as a stand. accessories. This exploration started with commercial selfie sticks [1] and then moved towards other low-cost accessories. This paper presents results of a user study that investigated impact of grip ring on mobile phone text entry performance in a stationary position. This is a continuation of a prior work that explored pointing accuracy on grip ring attached mobile phones [4]. Related Work Figure 4. Two and three-loop LAZY-HANDS grip. From LAZY- HANDS.com Figure 1. A user interacting with a grip Commonly available mobile phone grip rings enable ring attached mobile phone. From users to hold their devices safely. Users can slip their User Study Amazon. finger through the ring to keep their mobile phone We conducted a user study to investigate any potential firmly in hand (Figure 1). It also serves as a stand. impact of grip ring on text entry performance. Users can swivel the ring up to 360 degrees to set it at the ideal video watching angle (Figure 2). Kawabata et Apparatus al. [4] conducted a user study to investigate pointing We used a Google Pixel XL smartphone, 154.7 × 75.7 × accuracy on grip ring attached mobile phones. Results 8.5 mm, 168 g, at 534 ppi density during the study. We revealed that attaching grip ring improves pointing attached a Bunker Ring [9], 22.1 mm, at the center of accuracy for smaller targets and attaching the ring at the phone since this position yielded better pointing the middle yields relatively better results. accuracy in a prior study [4]. Bunker Ring is one of the most sold grip rings on Amazon [9]. We used WebTEM [2] to record text entry performance. Participants Seven participants, three female and four male, average age 27.4 years (SD = 3.8) participated in the study. All were right-handed and experienced mobile Figure 2. Grip ring swiveled in different users (over six years’ experience). Only one (female, degrees for different viewing angles. From Figure 3. PopSocket. From Minds Alive and Glik’s. 28 years) used a grip ring on her mobile device. Lynktec and Rakuten Global Market. Several alternatives to grip rings are available, most popularly PopSocket [7] and LAZY-HANDS [8]. The first Design is a collapsible grip and stand that is expanded to use We used a within-subjects design for the user study. (Figure 3) and collapsed to lay flat. The second is a two There were two conditions: with and without grip ring. The 7 Barcelona, Spain | September 3, 2018 MobileHCI 2018 Workshop on Socio-Technical Aspects of Text Entry conditions were counterbalanced. In each condition, participants transcribed 15 short English phrases [5] using WebTEM [2]. In summary the design was: 7 participants × 2 conditions × 15 phrases = 210 phrases, in total. Figure 5 shows a volunteer participating in the study. Procedure Upon arrival, we demonstrated the grip ring to all participants and allowed them to try it. We then started the study, where each participant transcribed 15 short English phrases from a set [5] using a smartphone with and without grip ring. WebTEM [2] displayed one random phrase from the set at a time and asked them to Figure 6. Average entry speed for the two examined transcribe it. Once done, participants had to press the conditions. Error bars represent ±1 standard error (SE). “Enter” key to see the next phrase. All participants Figure 5. A participant transcribing text Error Rate (ER) using a grip ring attached mobile phone. transcribed text in a seated position (Figure 5). They were instructed to hold the device in portrait position An ANOVA failed to identify a significant effect of grip with the dominant hand and then input with the thumb of ring on ER (F1,6 = 1.08, p > .05). On average ER the same hand. They used the default Android keyboard. without and with grip ring were 0.32% (SE = 0.09) and However, we disabled all predictive features, including the 0.46% (SE = 0.09), respectively. See Figure 7. prediction bar, auto-correction, capitalization, and custom dictionary, to eliminate a potential confound. Error correction was recommended, but was not forced [3]. WebTEM [2] logged all major text entry performance metrics [3]. At the end of the study, participants completed a short questionnaire about the grip ring. Results We used a repeated-measures ANOVA for all analysis. Words per Minute (WPM) An ANOVA failed to identify a significant effect of grip ring on entry speed (F1,6 = 0.05, p > .05). On average entry speed without and with grip ring were 25.56 WPM Figure 7. Average error rate for the two examined conditions. (SE = 1.09) and 25.59 WPM (SE = 1.08), respectively. Error bars represent ±1 standard error (SE). Figure 6 illustrates this. 8 Barcelona, Spain | September 3, 2018 MobileHCI 2018 Workshop on Socio-Technical Aspects of Text Entry Keystrokes per Character (KSPC) mobile phone usage and users’ interest in using grip An ANOVA failed to identify a significant effect of grip rings or similar devices. In other words, whether heavy ring on KSPC (F1,6 = 2.68, p > .05). On average KSPC mobile users are more likely to adapt to these devices without and with grip ring were 1.13 (SE = 0.03) and or not. Interestingly, we observed that different users 1.08 (SE = 0.03), respectively. Figure 8 illustrates this. use different fingers with grip rings. This could be due to different hand sizes. Yet, we cannot validate this due Qualitative Data to insufficient data. Most participant (57%, N = 4) did not feel that grip ring had any impact on entry speed or accuracy, while 29% Future Work (N = 2) felt it improved their entry speed and accuracy. In the future, we will include more participants in the The remaining 14% (N = 1) were neutral. Only 29% study. Since grip rings are likely to be more useful (N = 2) responded that they would consider using grip while walking, we will investigate whether it influences ring on their mobile devices, one of them was already text entry performance on the go. We will also explore Figure 8. Average keystrokes per character using a grip ring. 57% (N = 4) responded that they if handedness and different hand sizes impact grip ring for the two examined conditions. Error bars would not use grip rings, while the remaining 14% (N = usage, preference, and performance. Finally, we will signify ±1 standard error (SE). 1) were neutral. expand our investigation to various grip ring alternatives, such as PopSockets and LAZY-HANDS. Grip Finger We observed that participant exclusively used index Conclusion finger (71%, N = 5) or the ring finger (29%, N = 2) for We presented results of a study that suggested that the grip ring. This could be due differences in hand grip rings do not affect text entry performance in terms sizes. However, we do not have sufficient data to fully of speed, accuracy, and keystrokes per character. We investigate this. then reflected on future research directions based on the results and observations. The purpose of this work Discussion was to highlight the importance of classifying and Results revealed that there was no significant effect of evaluating low-cost mobile phone accessories. grip ring on text entry performance. Both conditions yielded comparable entry speed, accuracy, and Proposed Workshop Scenario keystrokes per character. However, since target In the workshop, we wish to discuss potential effects of selection is arguably more difficult in mobile settings, various low-cost mobile phone accessories on text entry such as while walking and commuting, grip ring may performance. Our intent is to highlight the importance benefit text entry on the go. of categorizing and evaluating these devices. We also wish to discuss why certain mobile accessories are Qualitative data showed that most participants were more popular in some countries than in the others, reluctant on using grip rings on their devices. It may be focusing on the sociotechnical aspects of these worthwhile to investigate if there is a link between accessories. 9 Barcelona, Spain | September 3, 2018 MobileHCI 2018 Workshop on Socio-Technical Aspects of Text Entry References 9. Bunker Ring. Retrieved June 13, 2018 from 1. Ahmed Sabbir Arif, Sunjun Kim, and Geehyuk Lee. http://www.bunkerring.com 2017. Usability of different types of commercial selfie sticks. In Proceedings of the 19th Biography International Conference on Human-Computer This section presents short biographies of the authors. Interaction with Mobile Devices and Services - MobileHCI ’17, 1–8. https://doi.org/10.1145/3098279.3098549 Monwen Shen is a MS student at the University of California, Merced. His research interests include 2. Ahmed Sabbir Arif and Ali Mazalek. 2016. WebTEM: embedded systems and Internet of things. a Web application to record text entry metrics. In Proceedings of the 2016 ACM on Interactive Surfaces and Spaces - ISS ’16, 415–420. Gulnar Rakhmetulla is a PhD student at the https://doi.org/10.1145/2992154.2996791 University of California, Merced. Her research focuses 3. Ahmed Sabbir Arif and Wolfgang Stuerzlinger. on text entry evaluation metrics and the design and 2009. Analysis of text entry performance metrics. development of accessible text entry techniques. In Proceedings of the IEEE Toronto International Conference - Science and Technology for Humanity Ahmed Sabbir Arif is an Assistant Professor at the - TIC-STH ’09, 100–105. University of California, Merced, where he leads the https://doi.org/10.1109/TIC-STH.2009.5444533 Human-Computer Interaction Group. A major thread of 4. Yuya Kawabata, Daisuke Komoriya, Yuki Kubo, his research focuses on smarter solutions for text entry Buntarou Shizuki, and Jiro Tanaka. 2016. Effects of holding ring attached to mobile devices on pointing and editing. He also models text entry performance and accuracy. In Human-Computer Interaction. designs text entry techniques for underrepresented Interaction Platforms and Techniques. 309–319. languages and user groups. https://doi.org/10.1007/978-3-319-39516-6_30 5. I. Scott MacKenzie and R. William Soukoreff. 2003. Phrase sets for evaluating text entry techniques. In CHI ’03 extended abstracts on Human factors in computer systems - CHI ’03, 754. https://doi.org/10.1145/765968.765971 6. Global Mobile Phone Accessories Market 2017- 2023. PR Newswire. Retrieved June 12, 2018 from https://www.prnewswire.com/news- releases/global-mobile-phone-accessories-market- 2017-2023-300483437.html 7. PopSockets. Retrieved June 13, 2018 from https://www.popsockets.com 8. LAZY-HANDS. Retrieved June 13, 2018 from https://www.lazy-hands.com 10