You may notice some images loading slow across the Open Collections website. Thank you for your patience as we rebuild the cache to make images load faster.

Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Comparing haptic application design communities : characterizing differences and similarities for future… Chun, Matthew (Jungho) 2020

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2020_may_chun_matthew.pdf [ 10.92MB ]
Metadata
JSON: 24-1.0389974.json
JSON-LD: 24-1.0389974-ld.json
RDF/XML (Pretty): 24-1.0389974-rdf.xml
RDF/JSON: 24-1.0389974-rdf.json
Turtle: 24-1.0389974-turtle.txt
N-Triples: 24-1.0389974-rdf-ntriples.txt
Original Record: 24-1.0389974-source.json
Full Text
24-1.0389974-fulltext.txt
Citation
24-1.0389974.ris

Full Text

Comparing Haptic Application Design Communities:Characterizing Differences and Similarities for FutureDesign Knowledge SharingbyMatthew (Jungho) ChunB.A., The University of British Columbia, 2016A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinThe Faculty of Graduate and Postdoctoral Studies(Computer Science)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)May 2020c© Matthew (Jungho) Chun, 2020The following individuals certify that they have read, and recommend to the Faculty ofGraduate and Postdoctoral Studies for acceptance, the thesis entitled:Comparing Haptic Application Design Communities: Characterizing Differences and Simi-larities for Future Design Knowledge Sharingsubmitted by Matthew (Jungho) Chun in partial fulfillment of the requirements for thedegree of Master of Science in Computer ScienceExamining Committee:Karon MacLean, Computer ScienceSupervisorTamara Munzner, Computer ScienceExamining Committee MemberiiAbstractHaptic technology has increasingly blended digital and physical world elements to createintuitive interactions in areas such as affective computing, VR/AR, video games, educationand various other domains.However, we posit that the emergence of best processes for designing impactful hap-tic applications has been hindered by a lack of shared understanding of the technical andconceptual design knowledge involved in developing meaningful haptic experiences.With over 27 years of diverse haptic literature, we have an opportunity to verify oursupposition by characterizing community design practices in their similarities/differenceswhich can be used to highlight areas of design expertise and gaps that other communitiescan help improve/complete. In future work, these characterizations can be further analyzedand integrated to help formulate effective haptic application design processes which couldlead towards new or improved haptic application experiences.We conducted a scoping literature review that provided initial characterizations of com-munity haptic application design practices, in order to lay the foundations for future cross-fertilization of design knowledge.iiiLay SummaryHaptic technology and its ability to recreate various types of physical sensations can en-able the physicality missing in applications such as affective computing and virtual realityimmersion.But the impact of haptic technology has been small. Its use in mainstream applicationshas been relegated towards simple vibrations in smartphones and video game controllers.This small impact could be due to a lack of coordination between notable application de-sign communities of the important technical and conceptual knowledge required to createmeaningful haptic applications.We used a scoping literature review to characterize, analyze, and compare notable hap-tic application design communities’ effective design practices regarding the conceptual andtechnical aspects of designing haptic applications. This was done in order to support futuredesign knowledge sharing that could serve to improve the overall process of haptic applicationdesign.ivPrefaceThis thesis was written based on the study approved by the UBC Behavioural ResearchEthics Board (certificate number H13-01620). This thesis represents original work thatprovides structured synthesis and analysis of multiple publications using a scoping literaturereview. Prof. Karon MacLean assisted in framing and editing sections of this thesis. SPIN(Sensory Perception and Interaction) lab members contributed feedback on various thesiscomponents such as the methodology and the overall motivations of the thesis.vTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiLay Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ixList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xAcknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1.1 Research Goal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1.2 Opportunity for Knowledge Sharing . . . . . . . . . . . . . . . . . . . 31.2 Research Questions and Methodological Approach . . . . . . . . . . . . . . . 31.2.1 Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.2.2 Methodological Approach . . . . . . . . . . . . . . . . . . . . . . . . 51.3 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Background: Communities That Do Haptic Design . . . . . . . . . . . . . 62.1 Haptics Community Background . . . . . . . . . . . . . . . . . . . . . . . . . 72.1.1 Origins of the Haptics Community . . . . . . . . . . . . . . . . . . . 72.1.2 Examples of Application Design in the Haptics Community . . . . . . 72.1.3 Hints of Haptic Application Design Practices . . . . . . . . . . . . . . 82.2 HCI Community Background . . . . . . . . . . . . . . . . . . . . . . . . . . 92.2.1 Rising Interest in Haptic Technology . . . . . . . . . . . . . . . . . . 92.2.2 Examples of Haptic Technology Usage in the HCI Community . . . . 9vi2.2.3 Hints of Haptic Application Design Practices . . . . . . . . . . . . . . 102.3 Reviews and Compilations of Haptic Application Design . . . . . . . . . . . 103 Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133.1 Focal Points and Research Questions . . . . . . . . . . . . . . . . . . . . . . 133.1.1 Focal Point 1 (Design Energy Distribution) . . . . . . . . . . . . . . . 163.1.2 Focal Point 2 (Purpose of Haptics in User Experiences) . . . . . . . . 173.1.3 Focal Point 3 (Design Methods Used to Create Haptic Applications) . 173.1.4 Focal Point 4 (Understanding Community Perspectives on Haptic Ap-plication Design) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183.2 Choosing Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183.2.1 Why a Literature Review? . . . . . . . . . . . . . . . . . . . . . . . . 183.2.2 Literature Review Overview . . . . . . . . . . . . . . . . . . . . . . . 193.2.3 Why a Scoping Review? . . . . . . . . . . . . . . . . . . . . . . . . . 193.3 Scoping Review Methodological Requirements . . . . . . . . . . . . . . . . . 203.4 Scope of Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213.4.1 Quantitative Data Clarity . . . . . . . . . . . . . . . . . . . . . . . . 213.4.2 Qualitative Data Clarity . . . . . . . . . . . . . . . . . . . . . . . . . 214 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224.1 Summary of our Scoping Review . . . . . . . . . . . . . . . . . . . . . . . . . 224.2 Research Space Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234.3 Stage 1. Determining Scope of Venues/Years . . . . . . . . . . . . . . . . . . 254.4 Stage 2. Identification of Possible Haptic Application Design Papers . . . . . 274.4.1 Step 2A. Keyword Extraction and List Creation . . . . . . . . . . . . 294.4.2 Step 2B. Keyword List Creation . . . . . . . . . . . . . . . . . . . . . 294.4.3 Step 2C. Using Suggestive Keywords to Identify Papers . . . . . . . . 324.5 Stage 3. Screening Identified Papers for Research Appropriateness . . . . . . 334.5.1 Step 3A. Title/Abstract Checking . . . . . . . . . . . . . . . . . . . . 354.5.2 Step 3B. Distribute Papers Among Haptic Experts . . . . . . . . . . 354.5.3 Step 3C. Haptic Expert Candidate Paper Inclusion Checking . . . . . 364.6 Stage 4. Analysis of Screened Papers . . . . . . . . . . . . . . . . . . . . . . 374.6.1 Step 4A. Final Paper Suitability Check . . . . . . . . . . . . . . . . . 404.6.2 Step 4B. Analysis of Expert Scores . . . . . . . . . . . . . . . . . . . 404.6.3 Step 4C. Qualitative Analysis By First Author . . . . . . . . . . . . . 404.6.4 Step 4D. Keyword Class Similarity Analysis . . . . . . . . . . . . . . 41vii5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445.1 Focal Point 1 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445.1.1 Focal Point 1 Results - By Community, a Broad View: HCI and Haptics 485.1.2 Focal Point 1 Results - By Year, a Historical View: 2014 and 2018 . . 505.1.3 Focal Point 1 Results – By Venue Size: Larger and Smaller . . . . . . 525.2 Focal Point 2 and Focal Point 3 Results . . . . . . . . . . . . . . . . . . . . . 545.2.1 Qualitative Code Definitions . . . . . . . . . . . . . . . . . . . . . . . 565.2.2 Focal Point 2 Results - Popular Haptic Experiences and Reasons forHaptic Technology Usage . . . . . . . . . . . . . . . . . . . . . . . . . 585.2.3 Focal Point 3 Results - Popular Design Methods Used to Create HapticApplications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595.3 Focal Point 4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625.3.1 Focal Point 4 - Similarity of Words to Describe Haptic ApplicationDesign . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635.3.2 Focal Point 4 - Common Words Used to Describe Haptic ApplicationDesign . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665.3.3 Focal Point 4 – Haptic Expert Rating Patterns . . . . . . . . . . . . 676 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706.1 Limitations of Our Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706.2 Focal Point 1 (Design Energy Distribution) . . . . . . . . . . . . . . . . . . . 726.3 Focal Point 2 (Purpose of Haptics in User Experiences) . . . . . . . . . . . . 736.4 Focal Point 3 (Design Methods Used to Create Haptic Applications) . . . . . 736.5 Focal Point 4 (Understanding Community Perspectives on Haptic ApplicationDesign) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746.6 Overall Impressions - Are the Communities Different or Similar to Each Other? 756.7 Future Work - Suggestions for Improving Our Study . . . . . . . . . . . . . . 767 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848.1 Future Work - Systematically Finding Appropriate Venues . . . . . . . . . . 848.2 Papers Used to Generate Focal Points . . . . . . . . . . . . . . . . . . . . . . 858.3 Papers Used in Methodology and Analysis . . . . . . . . . . . . . . . . . . . 878.4 Expert Ratings of Included and Excluded Papers . . . . . . . . . . . . . . . 94viiiList of Tables8.1 HCI community example papers used to inspire our focal points. . . . . . . . 868.2 Haptics community example papers used to inspire our focal points. . . . . . 878.3 The 22 excluded papers based on title/abstract checking by the first author. 898.4 The 4 HCI papers deemed not suitable for further analysis by our experts.This shows 4 out of 14 papers deemed not suitable for further analysis. . . . 908.5 The 10 Haptics papers deemed not suitable for further analysis by our experts.This shows 10 out of 14 papers deemed not suitable for further analysis. . . 918.6 The single HCI paper deemed not suitable for final analysis by the first author.This shows 1 out of 8 papers deemed not suitable for final analysis. . . . . . 918.7 The 7 Haptics papers deemed not suitable for final analysis by the first author.This shows 7 out of 8 papers deemed not suitable for final analysis. . . . . . 928.8 The final included 24 HCI community papers used in our analysis.This repre-sents 24 out of 38 papers used for the final analysis. . . . . . . . . . . . . . . 938.9 The final included 14 Haptics community papers used in our analysis. Thisrepresents 14 out of 38 papers used for the final analysis. . . . . . . . . . . . 94ixList of Figures1.1 The two communities of interest in our study. What areas of research amongthem describe haptic application design? . . . . . . . . . . . . . . . . . . . . 23.1 An illustration of our focal point formulation process. Outer boxes representinitial possible focal points. Inner boxes represent converged focal points basedon outer boxes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153.2 Our final focal points and associated research questions based on cohesivenesswith focal points, implications, and a clear means for insight. . . . . . . . . . 164.1 A high level overview of our scoping review stages. . . . . . . . . . . . . . . . 234.2 An overview of how our research space changed at each stage of our scoping re-view. CHI, UIST, ToH and HS are acronyms for HCI and Haptics publicationvenues which will be elaborated in Section 4.3. . . . . . . . . . . . . . . . . . 244.3 Stage 1. Our estimation of some notable venue year’s haptic application designpapers. A venue was considered for inclusion at this stage based on keywordsonly. Papers were not read at this point. . . . . . . . . . . . . . . . . . . . . 264.4 Stage 2. An overview of identifying possible haptic application design papers.Real keyword data is not given. Please refer to the results for keyword dataexamples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284.5 Stage 2. HCI-venue (CHI 2014, CHI 2018, UIST 2018) keywords thought toevoke haptic application design. This is the original paper keyword data. . . 304.6 Stage 2. Haptics-venue (ToH 2014, ToH 2018, HS 2018) keywords thought toevoke haptic application design. This is the original paper keyword data. . . 314.7 Stage 3. An overview of screening papers. Please refer to the results for hapticexpert rating data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344.8 Stage 3. Haptic expert background information. . . . . . . . . . . . . . . . . 364.9 Stage 3. The scoring system criteria used by our experts. . . . . . . . . . . . 37x4.10 Stage 4. An overview of our analysis process up to Step 4C. Real paper namesare not shown. The qualitative codes and visualizations are not based on realpaper data. Please refer to the results for analysis data. Please refer to Fig4.11 for Step 4D details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394.11 Stage 4. An overview of deriving keyword classes and how they were analyzedfor similarity. Real keyword data is not given. Please refer to the resultsfor real data examples. Different colours represent different types of venuepairings. Higher saturation indicates higher similarity. . . . . . . . . . . . . . 435.1 Data summary of our community design energy analysis. . . . . . . . . . . . 465.2 Distribution of haptic experts’ ratings of their approved papers from HCI andHaptics venues, for three aspects of haptic application design. This figureshows the distribution of ratings based on all of the HCI and Haptics venuepapers given to experts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495.3 Distribution of haptic experts’ ratings of their approved papers from HCI andHaptics venues, for three aspects of haptic application design. This figureshows the distribution of ratings based on 2014 and 2018 HCI and Hapticsvenue papers given to experts. . . . . . . . . . . . . . . . . . . . . . . . . . . 515.4 Distribution of haptic experts’ ratings of their approved papers from HCI andHaptics venues, for three aspects of haptic application design. This figureshows the distribution of ratings based on the big and small sample size HCIand Haptics venue papers given to experts. . . . . . . . . . . . . . . . . . . . 535.5 Data summary of our qualitative analysis. . . . . . . . . . . . . . . . . . . . 555.6 All of the codes and definitions used in our qualitative analysis. . . . . . . . 575.7 How each community emphasized certain user application categories sup-ported by haptic technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . 585.8 How each community emphasized specific reasons for designing with haptictechnology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595.9 How each community emphasized identified design practices in each area ofhaptic application design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605.10 How each community emphasized certain haptic hardware to create hapticapplications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615.11 The results of our keyword similarity analysis. Orange represents HCI compar-isons. Blue represents Haptics comparisons. Green represents cross-communitycomparisons. Higher saturation indicates higher similarity. Please refer to Fig4.11 for similarity value calculation details. . . . . . . . . . . . . . . . . . . . 64xi5.12 Keyword class appearances for all analyzed venue pairs. Positioning andcolour scheme intentionally resembles Fig 5.11 for consistent visual language.However, the saturation scheme is not implemented in this figure. . . . . . . 655.13 Popular keyword classes among the communities. Keyword classes are generalterms derived from original keyword data via analysis. . . . . . . . . . . . . 675.14 Each expert’s rating pattern for each haptic application design area. . . . . . 688.1 A possible approach to find new venues for analysis. The actual number ofhaptic application design papers found are used for illustration. . . . . . . . 858.2 Haptic expert ratings of each community’s venues on different areas of hapticapplication design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 968.3 Haptic expert ratings of the 2014 community venues on different areas ofhaptic application design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 988.4 Haptic expert ratings of the 2018 community venues on different areas ofhaptic application design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 998.5 Haptic expert ratings of the bigger sample community venues on differentareas of haptic application design. . . . . . . . . . . . . . . . . . . . . . . . . 1018.6 Haptic expert ratings of the smaller sample community venues on differentareas of haptic application design. . . . . . . . . . . . . . . . . . . . . . . . . 102xiiAcknowledgementsTo start, I would like to thank my supervisor, Karon MacLean, and all of my fellow SensoryPerception and Interaction (SPIN) lab members for their longtime support. All of theseindividuals imparted excellent advice that taught me how to be better in communicatingand framing my ideas, and to always ensure that the best approach was being taken in myresearch. More importantly, these individuals were a source of great motivation, and keptme going even when I hit some metaphorical walls. My research skills have no doubt beensharpened by all of SPIN, but I will aspire to grow further as a researcher to match theirimpressive qualities. I would also like to thank my second reader, Tamara Munzner, forgraciously offering her time and excellent insights to improve my thesis. I would like toalso thank the Designing for People (DFP) staff such as Jocelyn McKay who also providedexcellent advice/resources for my scoping literature review. Finally, I would like to thankmy family, friends, and fellow graduate students who have provided excellent morale supportand advice to keep me going through these years.xiiiChapter 1IntroductionHaptics, the field of technology and research of recreating physical sensations, could be themissing key in conveying the physicality of natural user applications such as virtual realityand affective computing.However, the technology’s impact on mainstream user application experiences can be con-sidered minimal. Mainstream haptic user applications have largely been relegated towardssimple vibrations in smartphones, wearable devices, and video game controllers.This minimal impact could be due to the unique challenges of haptic application designand a slow pace of developing best practices and support systems to address them. Wedefine haptic application design to be the rationale, descriptions, and methods of how haptictechnology is designed towards a targeted human activity.Fortunately, two notable communities have been involved in producing over 27 yearsof haptic research that may contain helpful knowledge to maximize the design of hapticapplication experiences. The goal of this thesis is to characterize the differences/similaritiesin design practice between these communities as a means of understanding whether therewill be benefit in establishing greater intermixing of their knowledge and methods.1.1 Motivation1.1.1 Research GoalAs Figure 1.1 indicates, we will codify two notable communities ostensibly covering differenttypes of haptic application design.HCI Community - HCI, shorthand for human computer interaction, is a relativelylarge research community focused on different perspectives of human interactions with com-putational technology, spanning a large diversity of methods, purposes and types of technolo-1gies. Haptics-based interaction design research may comprise a small but growing fractionof this community’s overall attention. In terms of scale, the largest HCI conference CHI(Conference on Human Factors in Computing Systems) featured 666 publications in 2018,the most recent year included in this study.Haptics Community - The Haptics community, true to its name, covers haptic technol-ogy related research, such as the engineering or psychophysical aspects of haptic technology.The notion of design within this community maybe different from the HCI community andmay cover only a small portion of the community’s research output. In terms of scale, thelargest Haptics conference HS (Haptics Symposium) featured 57 publications in 2018.Figure 1.1: The two communities of interest in our study. What areas of research amongthem describe haptic application design?Our research goal is to discover the different/similar community design practicesused to design haptic applications. Specifically, we wish to gain insights into designpractices regarding ...• Technical Design - These are details related to haptic technology and other relevantdetails needed to implement a haptic application.• Experience Design - These are details related to targeted user experiences that candrive the requirements for a haptic application design.• Design Methods - These are details regarding the specific methods used in thetechnical and experience design of haptic applications.We argue that each area is applicable to the haptic application design approaches of both2communities. We wish to understand the different/similar practices in these areas in orderto help lay the foundations for cross-fertilization of design knowledge.1.1.2 Opportunity for Knowledge SharingWe define cross-fertilization as the act of knowledge sharing between different researchcommunities in order to solve common problems or to create new innovative experiences.The benefits of cross-fertilization have been seen in many research domains.For example, Computer Scientists have been inspired by the Psychology and Statisticscommunities to create compelling artificial intelligence applications [1]. HCI researchershave adopted social science practices such as field data gathering techniques in order tocapture the reality of in-situ human behavior [2]. Haptics researchers have leveraged thepsychophysical and biomechanical knowledge of other domains to inform practical haptictechnology decisions [3].These examples show that cross-fertilization have shown tangible benefits in the under-standing and development of impactful experiences. Thus, we believe that such knowledgesharing could be helpful for HCI and Haptics researchers designing haptic applications.For example, if one community has faced problems in a specific design area, they couldadopt different design practices from a community that has solved these problems. If bothcommunities practice similar design methods, these can be compared to strengthen theoverall efficacy of the design methods.But before any potential cross-fertilization can be conducted, it is important to under-stand the current nature of haptic application design practices among the HCI and Hapticscommunities.1.2 Research Questions and Methodological Approach1.2.1 Research QuestionsThe cross-disciplinary nature of our research created a challenge in the formulation of appro-priate research questions. In particular, it was difficult to determine the common groundsthat could fairly compare the HCI and Haptics communities respective haptic applicationdesign approaches.Thus, to set the direction for our primary study, we conducted an informal analysis on aset of potential haptic application design papers (not based on our final methodology/sample)to help develop our research direction.3These papers were chosen through a variety of means including personal knowledge andreviewing the publications from well-known haptic research venues. More specifically, thesepapers came from an informal literature review that looked at notable community venueswithin our years of interest, covering different size scales and haptic knowledge contributions(eg. novel engineering to support new types of interactions). Please refer to Chapter 3(Approach) for further clarity on our informal analysis process.We derived the following focal points (concepts related to our research goals) and theirassociated research questions. These questions were chosen as they could produce interestingdata and implications for community haptic application design practices.Focal Point 1 - Design Energy DistributionRQ1(a) What are the distributions of effort, which we will call design energy, putforth by the communities into the relevant technical, user experience, and methoddetails of haptic application design?RQ1(b) Have design energy distributions been changing under different circum-stances?RQ1(b i) Different on a community level?RQ1(b ii) Different on different years?RQ1(b iii) Different on different sized venues?Focal Point 2 - Purpose of Haptics in User ExperiencesRQ2(a) Who are the typical users being targeted for haptic support?RQ2(b) What are the design reasons for using haptics to support these users?Focal Point 3 - Design Methods Used to Create Haptic ApplicationsRQ3 What are the typical techniques and considerations used by the communitiesin practicing the technical and user experience aspects of haptic application design?Focal Point 4 - Understanding Community Perspectives on Haptic Appli-cation DesignRQ4(a) How does each community self-report their haptic application design ac-tivities, through the lens of their publications’ keywords?RQ4(b) How do practitioners representing the communities view haptic applica-tion design practices? Do they interpret each community’s published technical, userexperience, and method details similarly or differently?41.2.2 Methodological ApproachOur chosen methodological approach was a scoping review. This type of literature reviewwas chosen as it provided a structure to synthesize representative community literature onhaptic application design practices. We will explain the full details of why a scoping reviewwas the most appropriate form of literature review in Chapter 3 (Approach). We will explainthe exact methodology of our scoping review in Chapter 4 (Methodology).Generally, our scoping review involved:• Selecting haptic application design papers from representative community venues (con-ferences/journals). We selected papers that provided a balanced set of technical, userexperience, and design method details used in haptic application design.• Analyzing papers in a comparative, iterative manner inspired by frameworks such asdesign thinking [4], and qualitative coding techniques such as open coding [5].1.3 Contributions1. Methodology: A demonstration of how a scoping literature review may be used tocompare design practices between defined samples (e.g. community or period of time).This methodology could be used as a basis for other comparisons.2. Characterization: A first analysis of applicable haptic application design practicesof two notable haptic research communities (HCI and Haptics), with objective anddescriptive data on areas of similarity and differences.3. Implications: We present preliminary implications and suggestions of how hapticapplication designers could use each community’s design knowledge for more effectivehaptic application design.5Chapter 2Background: Communities That DoHaptic DesignAn initial understanding of community haptic application design behavior, such as possibleexpertise areas and types of design advice given, can motivate the need to investigate whethercommunity design practices are different/similar on a larger scale.We will provide historical backgrounds and examples of design behavior that are poten-tially emblematic of each community’s haptic application design behaviour. We will alsodiscuss a few papers that has tackled the topic of haptic application design.As mentioned in Chapter 1 (Introduction), the literature references in this section werenot part of the final set of analyzed papers. These papers were chosen from an initial readingof literature informed by our inital exploration of conducting a cross-disciplinary literaturereview. We chose papers within the range of years/types of venues notable in overall size,and types of haptic design knowledge.However, these papers were not part of the actual set of papers used for analysis due tothe extensive time/energy required for full analysis. For example, we looked at papers fromCHI (ACM Conference on Human Factors in Computing) and UIST (ACM Symposium onUser Interface Software and Technology) 2016. From these types of venues, we read papersin which their titles, metadata keywords, and abstracts alluded towards providing designdetails in each area of haptic application design described in Chapter 1 (Introduction). Thepurpose of the reading these papers was to provide initial premises of how haptic applicationdesign practices may or may not be different among the communities.62.1 Haptics Community Background2.1.1 Origins of the Haptics CommunityMany different research groups have contributed to haptic research areas such as perceptionmodelling and technological engineering. One important group were Engineers develop-ing technology to recreate physical sensations inside virtual reality environments. In 1992,ASME (The American Society of Mechanical Engineers) formed a teleoperators and virtualenvironments journal. This was one of the first notable academic venues for haptics relatedresearch [6]. In 2008, this North American meeting changed its name to IEEE HapticsSymposium [7].Biomedical Engineers also contributed to the development of haptic technology that al-lowed for safe but realistic medical training. Venues such as the IEEE (Institute of Electricaland Electronics Engineers) Transactions on Biomedical Engineering is just one example ofseveral venues utilizing haptic technology [8].Engineering researchers were often focused on overcoming the mechanical challenges ofrecreating haptic sensations [9]. Engineering researchers soon realized the importance of un-derstanding haptic perceptual processes to create salient haptic sensations. Thus, Engineer-ing researchers began to collaborate with Psychophysicists in order to marry technologicaladvances with important haptic perceptual knowledge.This led to collaborative venues where the Engineers and Psychophysicists could sharetheir research findings with each other. For example, a conference called Eurohaptics wasfirst held in 2001 [10]. In 2005, the North American Haptics Symposium combined with Eu-rohaptics to form a new meeting called Worldhaptics, which has continued to run in alternateyears on different continents [11]. While the European and North American and more recentAsian communities vary somewhat, each are consistent in the emphasis of engineering andpsychophysical research rather than human computer interaction related research.The Haptics community has shown examples of adopting ideas from other researchgroups. Thus, it would be interesting to see whether the community’s haptic applicationdesign approach is comparable to that of a design-oriented community (HCI).2.1.2 Examples of Application Design in the Haptics CommunityIn one of a few notable works describing haptic application design in a cross-disciplinarymanner, MacLean et al. provided insights into designing with haptics by drawing uponinteraction design and haptic perception knowledge such as user goals, and haptic roles7within a multimodal interaction experience [12]. Several areas of applications were alsodescribed.One area of applications have designed haptics to act as a form of active feedback whereknowing persistent information was critical, such as presentation timing and posture cor-rection [12]. This form of active feedback has also been found helpful in reducing errorsin physically operated tasks such as surgical training and navigation. For example, Bae etal. developed a needle insertion system during biopsy operations where the haptic signalswere used to assist needle steering [13]. Girbes et al. developed a haptic feedback sys-tem that sought to reduce pedestrian accidents at low bus driving speeds [14]. Wang andKuchenbecker developed a haptic alert system that used a white cane that encoded distanceinformation via vibrotactile signals. These signals warned users with visual impairments ofupcoming low hanging objects [15].Other application areas have been used to support immersion in augmented media ex-periences such as virtual object interaction, and multimodal movie viewing experiences [12].Gabardi et al. developed a novel haptic thimble fingertip wearable device that utilized amovable vibration based actuator that could convey virtual object edges and surface tex-tures [16]. Lee et al. described the development of algorithms that could map a video’scamera motions and sound effects into appropriate motion chair movements and vibrationeffects respectively [17].2.1.3 Hints of Haptic Application Design PracticesSome examples from the Haptics community suggest that haptic application design relevantknowledge may come in the form of different design tools informed by other creative mediums.Inspired by music composition, Lee et al. created VibScoreEditor, which used the conceptof vibrotactile clefs for the creation of vibrotactile effects [18]. Danieau et al. describedhaptic effects informed by cinematic camera movements to enhance multimodal film viewingexperiences [19].There have been some movements towards the creation of design tools focused on sup-porting design activities/concepts such as iteration and design spaces.Similar to the idea of video or audio clips, Enriquez and MacLean proposed the notionof haptic icons, pre-programmed haptic effects that could be utilized in a modular fashionto support haptic effect editor programs [20]. Building off this modular concept, Schnei-der and MacLean developed Macaron, a web-based vibrotactile editor that allowed hapticdesigners to browse examples of vibrotactile effects, and to rapidly create new effects bysketching/refining/mixing properties from other effects [21].8In design space relevant research, Seifi et al. explored user expectations and preferencesof different visual organization schemas for customizing affective vibration effects [22]. In alater study, Seifi and MacLean developed a visualization tool called VibViz that organizeddiverse vibration effects into various facets that could be filtered and explored according toaffective and practical user considerations [23].2.2 HCI Community Background2.2.1 Rising Interest in Haptic TechnologyTo the best knowledge of the paper authors, one of which is an established researcher span-ning both HCI and Haptics communities, the HCI community has only recently began de-signing with haptic technology. This established researcher, Karon MacLean, has over 30years of haptic design related publications covering haptic applications, design tool support,conceptual design, affective computation, and technical contributions.From interviewing this researcher (personal communication, Feb 15, 2019), it was roughlyin 2012 when the HCI community began to show an interest in using haptic technology. Thiswas based on the number of haptic technology involved papers appearing in the flagshipHCI conference - CHI (ACM CHI Conference on Human Factors in Computing Systems). Inparticular, the novel interaction techniques area exhibited more haptic-related papers. Thisresearcher also noted the prevalence of haptic technology in other HCI relevant venues suchas UIST (ACM User Interface Software and Technology).Thus, it would be interesting to investigate this observation in a detailed manner. Thiscan help understand how a design-oriented research community designs with haptic technol-ogy compared to other communities.2.2.2 Examples of Haptic Technology Usage in the HCI Commu-nityIn MacLean’s overview of haptic design guidelines, another popular area of haptic applica-tions has been affective computation. Haptic technology was described as helpful for enablingemotional qualities in various types of interactions [12].Ueoka et al. investigated pressurized air vortexes as a potential way to affect user’semotional states for stress relief applications [24]. Allen et al. investigated how a breathinganimal-like robot that reacted to user touch gestures could be used to mitigate stress andanxiety [25].9Similar to the Haptics community, some examples have also suggested that the HCIcommunity is also interested in haptic information feedback applications. For example, Tamet al. developed a haptic wearable system that utilized signals to indicate timing informatonwhen giving oral presentations [26]. Pan et al. investigated how haptic feedback could beused as a form of bookmarking, to aid in quickly finding a point of interest inside audiomaterials [27].2.2.3 Hints of Haptic Application Design PracticesThe HCI community has provided some application design support tools to assist in appli-cation ideation. For example, tools have been developed to support haptic media captureand conversion, and taxonomies for exploring possible design directions.Minamizawa et al. created a toolkit called TECHTILE, an easy to use haptic mediacreation toolkit. The toolkit focused on capturing auditory information that could thenbe transferred into vibrotactile equivalent effects using microphones, voice-coil actuators,and signal amplifiers [28]. Haptipedia is a visualization tool used to help design new haptictechnology by exploring hardware metadata based on important device and interaction designattributes. These attributes have been derived from over 30 years of haptic literature [29].Hamam et al. surveyed relevant virtual reality and QoS (Quality of Service) literature inorder to create a taxonomy for evaluating the user experiences of haptic enabled virtualenvironments. The taxonomy considered technical, psychological, and physiological factorsinto a mathematical model for evaluating the overall quality of a haptic application [30].The community has also provided some guidelines on using psychophysical ideas for apractical interaction context. Pusch and Lecuyer took theories of pseudo-haptic perceptualprocesses and provided contextualized tips for creating convincing pseudo-haptic illusions[31]. Loffler et al. used past colour research to test various physical colour metaphors andtheir ability to improve the physicality of tangible interactions [32].2.3 Reviews and Compilations of Haptic ApplicationDesignInterestingly, there are many haptic literature reviews focused on very specific design topics.For example, many haptic literature reviews have focused on specific application areassuch as driving support [33], medical training tasks such as needle insertion [34] and reti-nal surgery [35]. These reviews often focused on discovering the types of study designs(evaluation metric, tasks), and haptic models used.10While these literature review papers are useful, they lack the cross-disciplinary naturethat is a focus of our work.Some other examples of existing haptic application design literature suggests that ap-plication design knowledge is often represented as taxonomies and guidelines based on psy-chophysical foundations.For example, Hale and Stanney developed an exhaustive set of psychophysically drivenguidelines for informing haptic designers of the perceptual capabilities of various hapticreceptors in the human body (eg. mechanoreceptor, kinesthetic receptor) and how certainhaptic features such as skin motion, muscle tension, texture, edge detections can be bestconveyed using haptic technological parameters (eg. force exertion, timing of stimuli, etc)[36]. Sjostrom provided practical design considerations for haptic applications supportingthe visually impaired. These considerations involved providing clear navigation referencepoints, a means to search or get an overview of objects available to feel, spacing of virtualphysical objects in an environment and more [37].To our best knowledge, very few works have focused on the topic of cross-disciplinaryhaptic application design practices.As mentioned before, MacLean et al. provided insights into designing with haptics bydrawing upon interaction design and haptic perception knowledge such as user goals, andhaptic roles within a multimodal interaction experience [12]. To the best of our knowledge,this work remains one of the most notable cross-disciplinary haptic application design works.The work closest to our research topic and literature review approach is Song et al.’sdata-driven content and semantic analysis literature review of popular haptic interactioncategories across disciplines such as Computer Science, Engineering, and Psychophysics [38].Unlike other types of literature reviews, Song et al. used a data-driven approach by scrapingand aggregating over 6000 haptic interaction papers to create illustrative visualizations ofpopular cross-community haptic interaction design categories. We would certainly like toadopt elements of this approach, as several of Song et al’s objectives aligns with our ownresearch goal. Specifically, there was a common point of comparing haptic interaction areasacross different disciplines.If we had to offer a critique of Song et al’s work, it is that unlike other types of literaturereviews, the authors stopped short of full paper reading that may have provided deeperinterpretations of their topic. It is standard for literature reviews to employ multiple readers,and paper inclusion criteria checks to ensure paper appropriateness for a given research topic[39].Additionally, we believe our research is different from Song et al.’s in that we are notjust discovering the popular categories of haptic interaction. What we also care to report11are the exact techniques, and design rationales that addresses both the technical and userexperience aspects of a haptic application.12Chapter 3ApproachThis chapter will provide clarity on the formulation of our focal points and their associatedresearch questions. We will describe why a scoping review was chosen as the most appropriatetype of literature review. We will describe decisions regarding the scope of our data analysis.3.1 Focal Points and Research QuestionsAs mentioned in Chapter 1 (Introduction), the cross-disciplinary nature of our research incomparing two different types of research communities presented a challenge. In particular,it was difficult to formulate research questions that could meaningfully be applicable for bothcommunities. In order to address this challenge, we generated focal points to help createmore specific answerable research questions impactful for both communities and that couldscale in analysis time/energy based on our small author team size.In order to help generate focal points, we conducted an informal preliminary analysisusing some haptic application design papers. In our case, focal points were cross-communityconcepts related to our research goal of discovering different/similar design practices acrossthe HCI and Haptics communities. Our analysis involved reading several papers that werethought to allude towards haptic application design application details as based on theirtitles, meta-data keywords, and abstract details. Please note that this analysis process wasa precursor to our formal paper search and analysis process that will be described in Chapter4 (Methodology).The papers used to generate focal points were based on personal knowledge and papersfrom well-known haptic research journals/conferences, such as from CHI 2016 (ACM CHIConference on Human Factors in Computing Systems) and ToH 2016 (Transactions on Hap-tics). These papers were not part of the final set of papers used for analysis. Please refer tothe Appendix for the example papers used in our focal point generation process.13Another important constraint of our research was to find a manageable number of pa-pers that could be deeply analyzed for understanding community haptic application designpractices. In order to determine the exact venues/years to use in our study, we estimatedthe number of papers we thought to be representative of haptic application design detailsbased upon a keyword metadata analysis of different possible venues. This keyword analysiswill be described in Stage 1 of Chapter 4 (Methodology). We used these keywords and theresulting number of associated papers to help decide which venues and years could providepotentially impactful papers for analysis. Additional clarity on our study’s scope can befound in Chapter 4 (Methodology).We brainstormed and iterated upon several focal points of concepts involved in the techni-cal, user experience and design methods of haptic applications. The initial ideas for possiblefocal points were inspired from our readings of notable community venue papers hintingabout haptic application design practices in regards to our areas of interest (technical, userexperience and design methods of haptic applications). After generating many focal points,we converged them into focal points that yielded research questions with interesting implica-tions on cross-community design practices. Focal points were also chosen if clear data sources(eg. paper metadata keywords) could be used. Our chosen focal points provided a mix ofquantitative and qualitative data. These could be used to both describe and summarize theoverall prevalence of certain design behaviours.It is important to note that this style of research, a cross-disciplinary literature review,was new for the author team. Thus, there was much refinement and iteration of our focalpoints and research questions throughout the study.Please refer to Figure 3.1, to see a visualization of our focal point formulation process.Please refer to Figure 3.2 for seeing the rationale of why we chose our final focal points andthe types of data we thought must be gathered from our literature review.14Figure 3.1: An illustration of our focal point formulation process. Outer boxes representinitial possible focal points. Inner boxes represent converged focal points based on outerboxes.Figure 3.2 shows our chosen focal points based on their implications on cross-communitydesign practices and whether a clean source of insight could be used to answer related researchquestions.15Figure 3.2: Our final focal points and associated research questions based on cohesivenesswith focal points, implications, and a clear means for insight.3.1.1 Focal Point 1 (Design Energy Distribution)In Chapter 2 (Background), there were suggestions that community backgrounds may havedirected the types of haptic application design activities being conducted. The Hapticscommunity seemed to be focused on engineering problems, while the HCI community wereusing haptic technology in a human centered design context.Whether these behaviours were occurring on a larger scale required further verification.Thus, Focal Point 1 was chosen to capture quantifiable measurements of community invest-ments into the technical or experience design areas of haptic application design.The results of this focal point could implicate the exact design areas that each communityfelt obligated to detail. Data for this focal point could come be quantified scores indicating16the degree of detail provided on haptic application design areas.In order to maintain objectivity, these scores can be provided by community representa-tive haptic experts. These scores could be collated in different ways to see if design energieswere changing at all in different years, or from different sized academic venues.3.1.2 Focal Point 2 (Purpose of Haptics in User Experiences)We were curious about the types of user applications that were supported by the communi-ties. This would be useful for communities designing a user application addressed by anothercommunity. The new community could study the design practices of a veteran communityin order to make effective haptic application design decisions.Complementary to user experiences, we thought it was important to understand the pur-ported design reasons of using haptic technology to support user experiences. For example,MacLean et al. described that haptics could be used to achieve interaction goals such asinformation notification or physical constraint/guidance [12]. It would be interesting to com-pare the community’s perspectives on the “best role” that haptic technology should take inan interaction. This could be useful for communities wishing to achieve a specific interactiongoal, in which the community with the most expertise could be consulted.Given these curiosities, Focal Point 2 was chosen to characterize the space of each com-munity’s supported user experience areas. We anticipated that this focal point would requirea form of qualitative analysis that would involve reading and synthesizing the most notableuser experience details into labels common among the communities.3.1.3 Focal Point 3 (Design Methods Used to Create Haptic Ap-plications)We felt it was important to involve focal points that detailed the specific types of designmethods used in the technical and experience design of haptic applications. For example, itwould be useful to know about the haptic hardware used to implement applications. Thus,Focal Point 3 was chosen to catalogue the actual methods used in haptic application design.This type of information would be useful for cross-fertilization, where communities couldadopt specific design methods to solve certain design problems. Like Focal Point 2, weanticipated that qualitative analysis would be used to synthesize the notable types of designmethods into labels that were common among the communities.173.1.4 Focal Point 4 (Understanding Community Perspectives onHaptic Application Design)It was mentioned in Chapter 2 (Background) that not all research within the HCI andHaptics communities would necessarily describe haptic application design.For example, the HCI community focuses on interaction design involving a wide rangeof technologies such as VR/AR, smartphones, and more [40]. The Haptics community mayfocus on other types of haptic technology topics such as overcoming mechanical challenges[9] or investigating perceptual processes involved in haptic sensations [36].Thus, given the diverse research topics within each community, it was possible that thelanguage used to describe haptic application design would be different for each community.Similar to Song et al’s metadata characterizations of haptic interaction concepts, we argue itis important to understand community words and how they can be used to characterize theirrespective perspectives on what concepts go into haptic application design [38]. Furthermore,the words can be used to indicate a degree of agreement, which we will define as a calculatedvalue of the proportion of shared keywords between communities.Focal Point 4 was chosen to capture whether the communities were similar in their wordsdescribing their perspectives on haptic application design. We also believed this focal pointcould be addressed by comparing different community experts and their respective viewson the similarities/differences regarding haptic application design. This can be done bycomparing Focal Point 1’s expert quantifications of the technical, experience design, anddesign method details of each community’s haptic application design papers.3.2 Choosing Methodology3.2.1 Why a Literature Review?Our research goal was to discover the different/similar community design practices usedto create haptic applications. This was to assess the potential for cross-fertilization, orknowledge sharing of design practices between the HCI and Haptics communities.This large-scale research goal necessitated a methodology that could find, analyze, andsynthesize large sources of information.Literature reviews appeared to be the most appropriate methodology, as they examinedliterature to synthesize general patterns of a field, or to identify gaps in knowledge that couldserve as the basis for future research [41].However, many literature review options existed, each with different methodological con-18straints and research purposes. Thus, it was important to understand the differences amongliterature reviews in order to choose the most appropriate review.3.2.2 Literature Review OverviewIn general, literature reviews utilize representative papers to help answer large-scale researchquestions in different ways.One class of literature review involves the summarization of large literature bodies tounderstand the current state or issues of specific research topics.Traditional/Critical reviews summarize literature to critique theories using commonlyreported methods/results [42]. Conceptual reviews summarize literature to help under-stand the issues of a specific research topic [42].Another class of literature review rigorously screens literature to find informative researchfor high stakes situations.Systematic reviews help make important decisions in risky domains such as governmentpolicy and medical treatment [43]. These reviews aim to collect strong evidence in anobjective, reproducible manner in order to make an optimal decision. Scoping reviews canhelp characterize the existing nature of a research topic that can highlight knowledge gapsaddressable by future research initiatives [43].In our study, scoping reviews were decided to be the most appropriate literature review.3.2.3 Why a Scoping Review?For our research, we believe that scoping reviews were the best choice given the methodolog-ical constraints/limitations posed by other types of literature reviews.Both traditional/critical and conceptual reviews do not necessitate a strict methodol-ogy for justifying paper selections. Thus, there is a risk of paper selection bias that mayundermine our research goal of discovering large-scale community design insights [42].Systematic reviews necessitate strict, standardized criteria to find informative papers forobjective decision making. Unlike the medical domain, we are not aware of any standardizedmeans to conduct haptic application design literature reviews in a cross-community com-parative manner. Systematic reviews requirement of standardized procedures, such as thePRISMA system used by the medical domain, precluded these reviews as a literature reviewoption [44].Different researchers have used scoping reviews to perform cross-disciplinary analysis ofdifferent disciplinary methodologies.19For example, Coemans and Hannes used a scoping review to understand how arts-basedmethodology of community inquiry on vulnerable populations was practiced across the socialscience and education research communities. Notably, Coemans and Hannes defined a clearsearch and analysis strategy to investigate different rationales for using arts-based methods,the genres of methods used, and the benefits/limitations of arts-based methods [39]. Kitsonet al. used a scoping review to understand how immersive interactive technologies, suchas virtual reality and augmented reality, were designed to support positive physiological oremotional change in users. This review addressed how different research disciplines such asHCI and Psychology designed immersive technologies [45].These examples demonstrate that scoping reviews are powerful tools that can objectivelysearch and analyze relevant literature, even literature across different research disciplines.This cemented our confidence in using scoping reviews as our main approach in discoveringthe similarities/differences of community haptic application design practices.3.3 Scoping Review Methodological RequirementsSection 3.2 demonstrated why scoping reviews were the best choice in understanding poten-tial differences/similarities in community haptic application design practices. However, theexact methodological steps were not specified.Fortunately, Arksey and O’Malley outlined a helpful framework for conducting scopingreviews, intended to help find and summarize relevant information from appropriate researchpapers [46]. The framework described stages such as:1. Identifying the research question2. Identifying relevant studies3. Study selection4. Charting the data5. Collating, summarizing, and reporting the resultsIn terms of identifying research questions, Section 3.1 described how we formulated re-search questions through a preliminary analysis of haptic application design papers. We alsodescribed the justification of our chosen focal points and the types of data/implications thatcould address community haptic application design practices.20Chapter 4 (Methodology) will address the remainder of Arksey and O’Malley’s frameworkstages. We will describe the details involved in identifying and selecting relevant studies, aswell as the processes involved in analyzing (charting, collating and summarizing) the results.It is also worth noting that Arksey and O’Malley described this framework as beingmore iterative, despite the linear order of stages presented. As we’ll describe in Chapter 4(Methodology) and Chapter 5 (Results), we can attest to the iteration needed to find faircomparison points of HCI and Haptics community design practices.3.4 Scope of Data AnalysisChapter 4 (Methodology) will describe the exact steps involved in finding and analyzinghaptic application papers. We wanted to clarify how resulting data would be handled in theanalysis described in Chapter 4 (Methodology).3.4.1 Quantitative Data ClarityFocal Point 1 and Focal Point 4 were expected to yield quantitative data. This data wasexpected to be the distribution of score categories (each category representing a degree ofdesign detail) provided by experts on different areas of haptic application design.Readers may be surprised to see that this data will not be analyzed using inferentialstatistical techniques such as ANOVA. Inferential statistics was not used, due to the smallauthor team size limiting the number of papers for effective inferential statistics. Inferen-tial statistics was also not suitable due to our study’s focus on characterizing rather thanconfirming community differences in design practices. Future work could potentially useinferential statistics to confirm the relevance of our discovered design practices.3.4.2 Qualitative Data ClarityFocal Point 2 and Focal Point 3 were expected to yield qualitative data. As Chapter 4(Methodology) will describe, we will use qualitative analysis techniques inspired by opencoding that will characterize general patterns of community design behavior into labeledcodes [5]. These codes are useful in succinctly describing general patterns of haptic applica-tion design behavior. Arksey and O’Malley also suggested that mechanisms like codes couldbe used to summarize and report the distribution of analyzed behavior among the papersused in a scoping review [46]. This would be useful for our study, due to the large-scalenature of our research goal benefiting from numerical styles of data reporting.21Chapter 4MethodologyThis chapter will describe the detailed steps of our scoping review. We will also report theexact number of papers considered at each stage of our scoping review, in order to illustratethe size of research considered by our work.Specifically, we will describe the methodological steps inspired by Arksey and O’Malley’sscoping review framework. Our stages involved determining the right scope of academicvenues/years appropriate for our study, identifying possible haptic application design papers,screening papers for study appropriateness, and analyzing screened papers to answer ourresearch questions.When necessary, we will present figures that will visually clarify the methodological stepstaken.4.1 Summary of our Scoping ReviewFigure 4.1 visually summarizes our scoping review’s stages and their respective approaches.As Arksey and O’Malley described in their framework, it is not uncommon to iterate betweenor on certain stages of a scoping review. Interestingly, we took much time to determine theright academic venues that should be the source of potential haptic application design papers[46]. This is indicated by the reverse up and down arrows in Figure 4.1 between Stages 1 and2. Additionally, we performed successive refinement of our focal points, eventually coalescinginto the final focal points reported in earlier sections.22Figure 4.1: A high level overview of our scoping review stages.Please refer to each stage’s respective section after this summary for finer details.4.2 Research Space SizeMedical scoping reviews tend to report numbers on the overall research space (paper num-bers) being considered and curated by a scoping review methodology [47].Due to the cross-disciplinary nature of our study, we also wanted to report the large scaleof research that was being addressed, in order to highlight the challenges of determining thedegree of haptic application design research occurring in the HCI and Haptics communities.23Figure 4.2 summarizes how our methodology affected the research space at each stage ofour scoping review.Figure 4.2: An overview of how our research space changed at each stage of our scopingreview. CHI, UIST, ToH and HS are acronyms for HCI and Haptics publication venueswhich will be elaborated in Section 4.3.244.3 Stage 1. Determining Scope of Venues/YearsCommon to most researchers, using existing knowledge of conferences, and other academicorganizations can be a viable method for determining sources of relevant studies [46]. Thiswas the approach used in determining the appropriate venues (conferences, journals) that webelieved to be representative of haptic application design practices based on our preliminaryfocal point generation process described in Chapter 3 (Approach). Please note that thesevenue acronyms will be used going forward.We decided that the HCI community would be best represented by:• ACM Conference on Human Factors in Computing (CHI)• ACM Symposium on User Interface Software and Technology (UIST)We decided that the Haptics community would be best represented by:• IEEE Haptics Symposium (HS)• IEEE Transactions on Haptics (ToH)To our best knowledge, this study is one of the first studies investigating research fromdifferent types of communities involved in haptic application design. The main author teamfaced limitations in the amount of data that could reasonably be searched and analyzed asindicated by Figure 4.2. Based on that figure, our choice of four venues yielded 1380 paperswithout any inclusion checking. This was a large number of papers that could not be easilyanalyzed. Thus, scope reduction steps that filtered towards papers with relevant details wasa necessity.In terms of years, we focused on 2014 and 2018 for detecting any possible design practicechanges. These two years were chosen as it provided a means to anchor recent designpractices (2018) with an interval of time large enough to show potential changes in anypractices (2014).Involving more years or larger gaps of time could show a larger change in design practices.Due to the small author team size, we required a gap of time that could still show potentialchange yet remained feasible for analysis.Hence, only 2014 and 2018 were chosen as they represented important start and endpoints of a time period thought to represent important developments in haptic applicationdesign. These anchor point years was also substantiated by our haptic expert author.Our choice of focusing upon 2014 and 2018 was further cemented by our estimation of thenumber of haptic application design papers of different venue years. These estimations were25based on a keyword analysis, that will be described in Stage 2 Step B of our methodology.Paper metadata keywords were used to help find potential haptic application design papers,but these papers were not read, unlike in Stage 2 Step B of our methodology.Figure 4.3 shows the estimations of haptic application papers during the years 2014, 2016,and 2018, compared with the original total number of papers published in a communityvenue.Figure 4.3: Stage 1. Our estimation of some notable venue year’s haptic application designpapers. A venue was considered for inclusion at this stage based on keywords only. Paperswere not read at this point.For CHI, we decided that 2014 and 2018 would be sufficient for our study, due to thesimilar estimations of haptic application papers in each year (average of 44 papers). Basedon these similar estimations, we decided that dropping CHI 2016 would assist in keeping amanageable sample set for analysis and not risk much in losing helpful papers.For UIST, we decided that 2018 would be sufficient for our study, due to Figure 4.3indicating that the years 2014 contributed only 7 potential haptic application design papers.26Additionally, we wished to maintain parity with our time span anchors of 2014 and 2018, thusonly 2018 was decided to be kept. Fig 4.3 also indicated that 2018 had as many potentialhaptic application design papers as 2016 (20 versus 17 respectively). Given these reasons,we felt that we wouldn’t lose much by dropping UIST 2016 from our scope of years.We took some time to carefully consider the venue year choices for our Haptics venues.While one may suspect that the HS conference should also use the years 2014 and 2018 tomaintain parity with our HCI CHI conference, we did not do so based on our estimationnumbers. For HS, Figure 4.3 indicated that 2014 had 49 potential haptic application designpapers, and that 2018 had 11 potential papers. This disparity in numbers, unlike CHI,indicated that comparing these two HS years maybe unfair. By comparison, ToH seemed toprovide more stable estimations, where Figure 4.3 showed that 2014 and 2018 could servean average of 25 papers consistently. Thus, given the similar estimations of anchors years(2014 and 2018), it was decided that ToH should involve two years for analysis, while HSshould keep 2018 in order to maintain parity with our UIST 2018 choice.4.4 Stage 2. Identification of Possible Haptic Applica-tion Design PapersScoping reviews often use a fixed set of search terms as an inclusion criteria (ie. applyingthe same terms to all databases searched) [45].However, in Focal Point 4 we observed the likelihood that each of our communities useddifferent keywords for activities related to haptic application design. We suspected that ourcommunities may differ from those more conventionally scanned in scoping reviews.Thus, we needed another way to identify papers using keywords that would be applicableto both communities. Figure 4.4 illustrates the actions involved in the identification ofpossible haptic application design papers.27Figure 4.4: Stage 2. An overview of identifying possible haptic application design papers. Real keyword data is not given.Please refer to the results for keyword data examples.284.4.1 Step 2A. Keyword Extraction and List CreationIn order to discover useful haptic application keywords applicable to the communities, weextracted metadata keywords from our chosen venue/year papers.In order to extract keywords, we used a critical Export CSV feature available in theACM Digital Library and IEEE Xplore databases that held our venue papers. This featureexported a parsable comma separated value (CSV) spreadsheet file that contained importantmetadata information such as paper titles, authors, and keywords. Some venues such asEuroHaptics had to be disqualified due to archival reasons that precluded metadata retrieval.Once the CSV files were obtained, we created R scripts to parse each venue’s keywordsinto lists that could be reviewed by the main author team.4.4.2 Step 2B. Keyword List CreationWe reviewed the venue keyword lists and chose keywords suggestive of haptic applicationdesign related concepts based on the backgrounds of the HCI and Haptics communities.To clarify, for each venue/year, keywords were chosen independently and specifically forthat particular venue/year. We used these venue/year specific keywords in our next step ofidentifying possible haptic application design papers.Figure 4.5 and Fig 4.6 shows the keywords chosen for the HCI and Haptics communitiesrespectively.29Figure 4.5: Stage 2. HCI-venue (CHI 2014, CHI 2018, UIST 2018) keywords thought toevoke haptic application design. This is the original paper keyword data.30Figure 4.6: Stage 2. Haptics-venue (ToH 2014, ToH 2018, HS 2018) keywords thought toevoke haptic application design. This is the original paper keyword data.31Our thought process for choosing certain types of HCI-venue keywords were as follows:• We tended to choose keywords suggestive of haptic technology such as “tactile display”.• This was due to the background of the HCI community being focused on interactionsupport for human activities.• Keywords related to human activities were not a priority, as it was likely the HCIcommunity’s papers would implicitly cover an aspect of human activity support.• Examples of unused keywords from the HCI community included “command selection”or “health information”. These were keywords that implicated a human activity, butdid not directly imply the presence of haptic technology.Our thought process for choosing certain types of Haptics-venue keywords were as follows:• We tended to choose keywords suggestive of human activities such as “motor rehabili-tation”.• This was due to the background of the Haptics community being focused on developinghaptic technology.• Keywords related to haptic technology were not a priority, as it was likely the Hapticscommunity’s papers would implicitly cover technology details.• Examples of unused keywords from the Haptics community included “closed loop con-trol” or “electrovibration”. These were keywords that implicated knowledge relevantto haptic technology, but did not directly imply the support of human activities.For this study, we decided to use venue/year specific keywords, as we did not fully un-derstand the general types of keywords that could be possibly repeated/shared among thedifferent community venues. Future work could investigate the nature of each community’skeywords such that a standardized list of keywords can be created to identify haptic appli-cation design papers reliably.4.4.3 Step 2C. Using Suggestive Keywords to Identify PapersWe used our chosen keywords to find their associated haptic application design papers. Thisresulted in a set of candidate papers that appeared promising in providing haptic applicationdesign details.32We consolidated any duplicated papers into a single paper for consideration, as it was oc-casionally found that different keywords resulted in the same paper being identified. Search-ing with the keywords and consolidating duplicates led to 82 candidate papers (1380 paperswent into Stage 1 and Stage 1 dropped 1298 papers).4.5 Stage 3. Screening Identified Papers for ResearchAppropriatenessScoping reviews can require an inclusion criteria to screen out papers that does not meet aset of research objectives [39]. In order to conduct paper screening, two steps are commonlyused – title/abstract checking and multiple author checking of papers [46] [39] [45].Please refer to Figure 4.7 to see how these screening steps were applied.33Figure 4.7: Stage 3. An overview of screening papers. Please refer to the results for haptic expert rating data.344.5.1 Step 3A. Title/Abstract CheckingThe first author read the titles, abstracts and if necessary, the body of Stage 2 candidatepapers.This was required in order to verify whether the candidate papers were indicative of ourhaptic application design areas of interest. For a paper to be considered appropriate, itmust have described a balanced set of details pertaining to the technical design, experiencedesign, and design method details of haptic applications. If any one of these areas were notaddressed, the paper was screened out.This step led to 60 candidate papers considered more promising of haptic applicationdesign details (82 papers went into this Stage 3 step and this Stage 3 step dropped 22papers).4.5.2 Step 3B. Distribute Papers Among Haptic ExpertsAs seen in Figure 4.7 in order to avoid paper selection bias, we took Step 3A’s candidate paperset of 60 papers and consulted the help of three haptic experts. These experts reflected thedifferent communities of interest in order to ensure a fair voting process of screening hapticapplication design papers.We defined haptic experts to be published academics that were studying haptic researchunder a design perspective. These perspectives could be conceptual or technical in nature.Please refer to Figure 4.8 below to see our expert backgrounds.As we will describe in the next section below, each expert was given the same 60 can-didate papers from Step 3A to fully read and analyze for a screening task (Paper InclusionChecking).35Figure 4.8: Stage 3. Haptic expert background information.4.5.3 Step 3C. Haptic Expert Candidate Paper Inclusion Check-ingBased on our personal knowledge of academics that have studied haptic research from dif-ferent community backgrounds, we personally contacted interested academics as describedin Figure 4.8. Each academic expert was asked to perform a paper inclusion checking task.Each expert was tasked to independently read the 60 candidate papers from Step 3A. Eachexpert used the same 3-point scoring system that quantified the different levels of detailsgiven to each area of haptic application design (technical design, experience design, designmethods). Each of these areas were given a score out of 3 points, and the subsequent talliedscore would be used to indicate the overall level of haptic application design detail in a paper.The full details of this scoring system can be seen in Figure 4.9 below.We ensured paid compensation for time committed to the task. We estimated that the36task would take 3 to 4 hours to complete based on our piloting of the task. Each expertreported taking 4 hours to complete the task. Each expert was paid 40 Canadian dollars perhour.Figure 4.9: Stage 3. The scoring system criteria used by our experts.Only papers that were given passing scores of over 50% by at least 2 out of 3 experts wereconsidered appropriate for final analysis. This stage ended with 46 papers deemed suitablefor further analysis (60 papers went into this Stage 3 step and this Stage 3 step dropped 14papers).The expert scores were also used to address Focal Point 1 and Focal Point 4. Thiswas to characterize the design energies of different communities on different areas of hapticapplication design. This was also to understand how the different community experts weresimilar/different in their interpretations of haptic application design details. These will bediscussed further in Chapter 5 (Results).4.6 Stage 4. Analysis of Screened PapersOur focal points and their research questions suggested several types of data to be analyzed,ranging from quantitative to qualitative in nature. Thus, it was important to consider themanner of presentation of the paper data as it was being analyzed. We detail how we37presented our data below. Figure 4.10 describes our analysis process.38Figure 4.10: Stage 4. An overview of our analysis process up to Step 4C. Real paper names are not shown. The qualitativecodes and visualizations are not based on real paper data. Please refer to the results for analysis data. Please refer to Fig 4.11for Step 4D details.394.6.1 Step 4A. Final Paper Suitability CheckThe first author performed a final suitability check of the Stage 3 papers. This involved a fullreading of each of the Stage 3 papers. It was found that some papers had to be discountedfor further analysis, as some did not meet our inclusion criteria of providing a balanced setof haptic application design details.For example, some papers were haptic perceptual studies that did not have direct impli-cations on the experience design of haptic applications. Some papers were found to detailtechnical models/algorithms that were implicated but not informed by human activities.This reflects a limitation of our inclusion criteria provided to the haptic experts in Stage 3.After the final suitability check, 38 papers were left for final analysis (46 papers wentinto Stage 4 and Stage 4 dropped 8 papers).4.6.2 Step 4B. Analysis of Expert ScoresThis analysis addressed Focal Points 1 and 4, where we sought to understand the distribu-tions of design energies among the communities and whether community practitioners (ourexperts) viewed haptic application design differently from each other.We considered the quantitative expert scores of Stage 2 to be a good basis for addressingthese focal points. The expert scores indicated levels of design detail and could be visu-alized as distributions to identify the design areas of community focus and areas of expertagreement/disagreement.In order to address our focal points, the first author collated and visualized the distribu-tions of expert scores in different ways. To address Focal Point 1, the first author collated allthe expert scores into a collective distribution to highlight the consensus of how communitydesign energies were being distributed. To address Focal Point 4, the first author createddifferent distributions of each expert’s scoring behavior on different haptic application designareas.Please refer to Chapter 5 (Results) to see how we visualized the expert scores.4.6.3 Step 4C. Qualitative Analysis By First AuthorThis analysis addressed Focal Points 2 and 3, where we wished to specify the design methodsand typical user experience areas/reasons of designing with haptic technology.These focal points required qualitative information, thus we employed a qualitative codingapproach inspired by opening coding [5]. This involved deeply reading and re-reading the38 papers deemed suitable for final analysis in a comparative manner.40This goal of this process was to arrive at a set of codes that were simple words/phrasesthat would be reflective of a haptic application design behavior. For example, codes coulddescribe the types of hardware being used among the communities. The creation of thesecodes involved much iteration and clustering of other smaller codes into more representativelarger categories. This also involved looking back towards our focal points and researchquestions, and iterating upon them in order to find cohesive categories of focal points thatbest represented our code data in light of interesting implications for our communities. Pleaserefer to Chapter 5 (Results) to see the resulting codes.Once the codes were stabilized, we collated the codes to visualize distributions of com-munity user experience areas, the design methods used in the technical design, experiencedesign, design methods involved in haptic applications, and the purported design reasons todesign with haptic technology.4.6.4 Step 4D. Keyword Class Similarity AnalysisThis analysis addressed another important aspect of Focal Point 4, in which we sought tounderstand how the communities described haptic application design based on the keywordsbeing used.Due to the nature of keywords being different from each community, we opted to con-sider the class of keyword rather than the keyword itself in order to allow for similaritycomparisons. To be clear, keyword classes were independently and specifically made for eachvenue/year in our study. The keyword classes from each venue/year were then compared tosee if any classes were repeated among certain pairings of venues/years. These mechanismswere based on a human judgement process, and may have been prone to imperfections.The mechanisms we used to derive keyword classes were:• Typically from long keyword phrases, we selected root word as the keyword class whichwas most relevant to our purpose.• If different parts of a keyword phrase seemed indicative of a haptic application design,then each word of the phrase became a keyword class.• If the entirety of a keyword phrase signaled a unique perspective on haptic applicationdesign, then the whole phrase was used as a class.• Keyword classes were based on our interpretations of words that implied haptic appli-cation design from the perspective of each community.We give examples of how these mechanisms played out in the different venue groups.41For HCI venues:• Keyword classes tended to be haptic technology-related, given that the community hasbeen known to perform interaction design using non-haptic technologies.• For example, the keywords “haptic interface”,“haptic display”,“tactile display” weredistilled into the keyword classes of “haptic” and “tactile” to capture the technologicalaspect of the keywords.• In other cases, a keyword such as “tactile drawing” would be distilled into the 2 separatekeyword classes of “tactile” and “drawing” as these parts indicated a technology andan activity respectively.For Haptics venues:• Keyword classes tended to be human activity-based, given that the community hasbeen known to perform technological development not necessarily tied to any humanactivity.• For example, the keywords “driving skill”,“driving assistance”, “adaptive control”, and“motion control” would be distilled into the keyword classes of “driving” and “control”in order to capture the interaction aspects of these keywords.• In other situations, the entire original keyword was used as a keyword class. For ex-ample, the keywords “design parameter” and “design recommendations”, were decidedto be used as separate keyword classes despite the common presence of the “design”word among them.• Our thought process for the above keyword classes was that “design parameters” indi-cated knowledge contributions of a technical implementation, while “design recommen-dations” could involve guidelines/advice for approaching an haptic application designprocess. Thus, these sorts of keywords were kept as separate keyword classes.Thus, our mechanisms for deriving keyword classes generated different types of classesfor comparison. Please refer to Chapter 5 (Results) for examples of keyword classes.42Figure 4.11: Stage 4. An overview of deriving keyword classes and how they were analyzed for similarity. Real keyword datais not given. Please refer to the results for real data examples. Different colours represent different types of venue pairings.Higher saturation indicates higher similarity.43Chapter 5ResultsThis chapter will describe the results of our scoping review. The ordering of the results willaddress each of our focal points and associated research questions described in Chapter 1(Introduction) and Chapter 3 (Approach).The results will be presented primarily as visualizations. We will interpret each visualiza-tion by highlighting the relevant focal points, research questions, and involved data requiredto interpret each visualization. These details will be preparatory to developing insights thatwill be detailed in Chapter 6 (Discussion).In general, we will present visualized results that will compare community differences andsimilarities on haptic application design details such as the distributions of design energy,the design methods used, types of experiences/design reasons for using haptic technology,and how the communities described and viewed each other’s approach to design.5.1 Focal Point 1 ResultsAs a reminder, Focal Point 1 was defined as “Design Energy Distribution” where there was acuriosity in determining the differences/similarities of quantifiable design efforts put forth bythe communities into the technical or experience design aspects of haptic application design.Additionally, we were curious whether design energies were changing at all in different years,or from different sized academic venues, which could imply changes in haptic applicationdesign approaches under different contexts.Focal Point 1’s research questions were the following:RQ1(a) What are the distributions of effort, which we will call design energy, put forthby the communities into the relevant technical, user experience, and method details ofhaptic application design?44RQ1(b) Have design energy distributions been changing under different circumstances?RQ1(b i) Different on a community level?RQ1(b ii) Different on different years?RQ1(b iii) Different on different sized venues?Chapter 4 (Methodology) elaborated upon the consultation of different community hapticexperts, and their usage of a scoring system aimed to characterize how each community wasdetailing haptic application design. Please refer to Figure 4.8 for a reminder of our expertsand their qualifications for this research.The results of this section will cover how our experts collectively characterized differentcircumstances of design energy distributions, based upon our research questions in Chapter1 (Introduction).We will visualize the expert scores as visualized distributions by community, by differentyears, and by venue size. These address our research questions RQ1 (b i) to RQ1 (b iii)respectively.Figure 5.1 illustrates the background of our Focal Point 1 results, reminding readers ofthe data source used by haptic experts to determine different distributions of design energiesamong the communities. The figure also provides a map of the different circumstances ofdesign energy distributions that will be presented below.45Figure 5.1: Data summary of our community design energy analysis.How to Read Focal Point 1 Visualizations - For each visualization seen in thissection below, please note the following:1. Paper count (the number of papers that fell under a specific rating category) are46reported as the y-axis of the figures described in Fig 5.2, 5.3, and 5.4.2. Different colours indicate different communities.3. Trendlines indicate the general pattern (shape) indicated by the distribution.4. Visualizations report on the 46 expert verified (included) papers based on the outputof Stage 3 Step 3B.5. Score categories represent differences in the levels of detail provided by a community,where higher score categories indicate higher levels of details being provided.6. Each visualization will use different sample groups to represent specific circumstancesof RQ1 (b i) to RQ1 (b iii). The visualization header will specify which sample groupsare involved.7. For each visualization, the relevant sample sizes are reported in the top portion. Theserepresent the sum of count of papers approved by each expert for a particular data view(e.g. “By Community” would consider all venues/years). All experts worked from thesame pool of 60 candidate papers described in Step 3A of our methodology. However,each view covered different pools of community papers based on these 60 candidatepapers (e.g. 29 HCI papers, 31 Haptics papers for the community view). Additionally,each expert had discretion over how many papers they approved, and we captured thisdiversity by summing their individual approval counts. For example, in Fig 5.2, thefinal sample size of 67 HCI papers was due to the summation of 21, 24, and 22 approvedpapers from Experts 1 to 3 respectively. In other words, each expert approved differentamount of papers from the pool of 29 HCI papers available in the community view.Each visualization will denote the pool sizes, and other relevant sample information inthe top portion of the figure.8. Percentages Are Not Reported. These results report paper counts, not as per-centages. We wanted to emphasize any possible disparities/gaps between the heightsof different score categories which are computed on different bases. Thus, a percentagewould distort these true differences.We also visualized expert ratings to show the distributions of included and excludedpapers. This was to help identify what details pushed a community’s set of papers intobeing considered appropriate for haptic application design analysis. These distributions canbe seen in the Appendix section.47Interpretation Example - Please refer back to Figure 4.9 for a reminder of the hapticexpert’s scoring system used to characterize the levels of details given in each area of hapticapplication design. Suppose an HCI community set of venues (eg. CHI 2014/2018) displayedhigher occurrences of the Score 3 category within technical design compared to the Hapticscommunity venues. This could be interpreted as the HCI community prioritizing technicalhaptic details.If the HCI community displayed high occurrences of lower score categories (eg. 0 and 1)in design areas such as experience design in the excluded paper score distribution, this couldbe interpreted as the community placing lower value in providing such details. Figures thatdisplay excluded paper score details can be seen in the Appendix chapter.5.1.1 Focal Point 1 Results - By Community, a Broad View: HCIand HapticsAt the top of the following figures, “Papers Rated By Experts” indicates the total papersgiven to the haptic experts for rating, for each community’s venues (total 60). “Total Ap-provals” sums the papers approved by each haptic expert, for a combined total greater thanthose that were rated. Note that low ratings can appear (at a lower expected frequency thanthe higher ratings) because a given expert could provide at least one lower rating but stillapprove that paper. If a given expert did not approve a paper, their ratings for that papersare not included here.Figure 5.2 represents how the communities behaved differently/similarly in haptic ap-plication design. The figure grouped all HCI venues against all Haptics venues used in ourstudy.48Figure 5.2: Distribution of haptic experts’ ratings of their approved papers from HCI andHaptics venues, for three aspects of haptic application design. This figure shows the distri-bution of ratings based on all of the HCI and Haptics venue papers given to experts.49Overall Impressions - Both communities were rated similarly on almost all areas ofhaptic application design. This can be seen in the similarly shaped trendlines among thedifferent haptic application design areas in the top section of Figure 5.2.Experience design was an exception, where the HCI community had higher occurrencesof Score 3 rating category (30) over the Haptics community (13).Thus, we are left with an impression that the HCI community may prioritize experiencedesign details slightly more than the Haptics community.5.1.2 Focal Point 1 Results - By Year, a Historical View: 2014and 2018Please note that the venues analyzed were only CHI 2014/2018 and ToH 2014/2018. Therewere no 2014 iterations for UIST and HS in which a comparison could be made. Please referto Chapter 4 (Methodology) for justifications on this matter.Figure 5.3 represents both a baseline view (2014) in which future venues (2018) can beused to detect any changes in haptic application design patterns.50Figure 5.3: Distribution of haptic experts’ ratings of their approved papers from HCI andHaptics venues, for three aspects of haptic application design. This figure shows the distri-bution of ratings based on 2014 and 2018 HCI and Haptics venue papers given to experts.Overall Impressions - In 2014, both communities were rated similarly on all areas ofhaptic application design. This was even the case for experience design ratings.By contrast in 2018, a growth in the HCI community could be seen in all areas of hapticapplication design. This can be seen as higher occurrences of the Score 3 rating category. For51example, in 2018‘s technical design, HCI showed 14 occurrences over Haptics’ 4 occurrences.In 2018’s experience design, HCI showed 17 occurrences over Haptics’ 3 occurrences. In2018’s design methods, HCI showed 11 occurrences over Haptics’ 4 occurrences.2014 indicated the beginning of a trend in which both communities seemed comparable inall areas of haptic application design. However in 2018, both communities exhibited differenttypes of growth in different areas of haptic application design.The HCI community showed the biggest change among the higher score categories ofdifferent haptic application design areas.5.1.3 Focal Point 1 Results – By Venue Size: Larger and SmallerFigure 5.4 represents the community design patterns of the bigger sized venues (based onsample size). This view can be compared against the smaller sized venues in order to denoteany differences/similarities in design patterns.52Figure 5.4: Distribution of haptic experts’ ratings of their approved papers from HCI andHaptics venues, for three aspects of haptic application design. This figure shows the dis-tribution of ratings based on the big and small sample size HCI and Haptics venue papersgiven to experts.53Overall Impressions - For the big venues, both communities were rated similarly onall areas of haptic application design. Similar to the community view, the HCI communityprioritized experience design details. This is indicated by the higher occurrences of the Score3 rating category (23) over the Haptics community’s occurrences (12).For the small venues, the HCI community had high occurrences of the Score 3 rating cat-egory in technical design (7), experience design (7) and design methods (6). By comparison,the Haptics community had lower occurrences of the Score 3 rating category in technicaldesign (4), experience design (1) and design methods (0).The big venue patterns suggested that the HCI community prioritized experience designdetails. The smaller community venues behaved somewhat differently than the bigger venues.The HCI community seemed to provide more details in all haptic application design areas.5.2 Focal Point 2 and Focal Point 3 ResultsFocal Point 2 was defined as “Purpose of Haptics in User Experience” where we wanted tounderstand the different community purposes of using haptic technology to support specificuser experiences. Focal Point 3 was focused on“Design Methods Used to Create Haptic Ap-plication” where we wished to characterize the specific design methods used in the technicaland experience design areas of haptic applications.Focal Point 2’s research questions consisted of the following:RQ2(a) Who are the typical users being targeted for haptic support?RQ2(b) What are the design reasons for using haptics to support these users?Focal Point 3’s research questions consisted of the following:RQ3 What are the typical techniques and considerations used by the communities inpracticing the technical and user experience aspects of haptic application design?Chapter 4 (Methodology) described our qualitative analysis process in deriving codes todescribe community design practices, types of user experiences, and associated reasons forthe usage of haptic technology.The results of this section will visualize the distributions of our qualitative codes. Figure5.5 illustrates the background of our Focal Point 2 and 3 results, reminding readers ofthe data sources involved in our qualitative analysis. The figure also provides a map ofthe different aspects (RQ2a/b and RQ3) of haptic application design practices and userapplication considerations being visualized.54Figure 5.5: Data summary of our qualitative analysis.How to Read Focal Point 2/3 Visualizations - For each distribution visualizationseen in this section below, please note the following:1. The percentage of papers exhibiting a qualitative code is reported. Wedecided to use percentages as we wanted to compare how each qualitative code was usedamong the differently sized sets of community papers. The uneven community sampleswere a byproduct of our screening methods described in Chapter 4 (Methodology).552. Percentages when added up for a community may exceed 100%. This is dueto papers having multiple applicable qualitative codes.3. Different colours indicate different communities.4. We provide initial impressions of the data, without jumping into full discussion.5.2.1 Qualitative Code DefinitionsAs described in Chapter 4 (Methodology), the generation of our qualitative codes involvedmuch iteration and clustering in order to find the most succinct and applicable codes acrossboth community’s set of verified haptic application design papers.Figure 5.6 details the codes and their definitions displayed in our distributions below.56Figure 5.6: All of the codes and definitions used in our qualitative analysis.575.2.2 Focal Point 2 Results - Popular Haptic Experiences and Rea-sons for Haptic Technology UsagePopular Haptic ExperiencesFigure 5.7 shows the results of our qualitative analysis of the popular haptic user experiencescategories supported by the communities.Figure 5.7: How each community emphasized certain user application categories supportedby haptic technology.Overall, there is an impression that the communities had different preferences in thetypes of user applications being designed.The HCI community supported applications for social communication (13%), visual im-pairment (13%), and professional physical user control (13%).Meanwhile, the Haptics community supported medical related applications such as med-58ical professional support (36%) and physical rehabilitation (36%).Only driving applications shared any occurrences between the communities. The HCIcommunity exhibited 13% of papers detailing driving support, while the Haptics communityexhibited 7%.Popular Reasons for Haptic Technology UsageFigure 5.8 shows the results of our qualitative analysis on the popular reasons for haptictechnology usage in supporting user experiences.Figure 5.8: How each community emphasized specific reasons for designing with haptictechnology.The types of designed user experiences was notably different among the communities.The HCI community emphasized haptic technology as a form of information display(75%).The Haptics community emphasized haptic technology as a means to improve physicalcontrol (71%).5.2.3 Focal Point 3 Results - Popular Design Methods Used toCreate Haptic ApplicationsFigure 5.9 shows the results of our qualitative analysis on the popular design practices seen indifferent areas of haptic application design - technical design, experience design, and designmethods.59Figure 5.9: How each community emphasized identified design practices in each area ofhaptic application design.Overall, there is an impression that the communities emphasized design practices verydifferently.For instance, the HCI community reported general system architecture details (79%).Meanwhile, the Haptics community reported a variety of technical design details such asalgorithms (86%) and stimuli parameters (50%). To clarify, general system architecturedetails described technical details that generally listed the names of the hardware/softwarebeing used, but did not specify how these were used to implement haptic experiences. Mean-while, algorithms and stimuli parameters provided the descriptive step by step actions thathardware/software should take to create a target haptic sensation.As Figure 5.7 reported, the types of designed user experiences was notably differentamong the communities. In terms of the design methods being reported, the Haptics com-munity typically did not report any method details (71%). The HCI community reported60use cases (50%) and design spaces for haptic applications (50%).Popular Types of Haptic HardwareFigure 5.10 shows the results of our qualitative analysis of the prevalence of popular haptictechnologies used in the technical design of haptic applications.Figure 5.10: How each community emphasized certain haptic hardware to create hapticapplications.Overall, there is an impression for different preferences of haptic hardware among thecommunities.The HCI community showed a strong preference for tactile display technologies (46%).The Haptics community exhibited a strong preference for force feedback displays (64%).Interestingly, both communities showed strong interest in vibrotactile display technolo-gies. The HCI community showed a preference of 21%, while the Haptics community had apreference of 36%.615.3 Focal Point 4 ResultsFocal Point 4 was focused on “Understanding Community Perspectives on Haptic Appli-cation Design” where we we wished to uncover the differences or similarities in how thecommunities described haptic application design. These could be uncovered by looking atthe language or words used by each community on describing haptic application design con-cepts. Additionally, we recruited the help of three community representative haptic expertsand collected their ratings of the different levels of details provided by the communities onhaptic application design areas.Focal Point 4’s research questions were the following:RQ4(a) How does each community self-report their haptic application design activities,through the lens of their publications’ keywords?RQ4(b) How do practitioners representing the communities view haptic applicationdesign practices? Do they interpret each community’s published technical, user expe-rience, and method details similarly or differently?Chapter 4 (Methodology) described how we analyzed the different community keywordsused to describe haptic application design papers. It was mentioned that similarity met-rics, calculated by the proportion of the shared keyword classes between two venues, werecreated to help understand the degree of similarity in community descriptions of haptic ap-plication design. We also described how the experts used a scoring system to independentlycharacterize how each community was detailing different areas of haptic application design.The results of this section will cover the different similarity values of all pairings of ourcommunity venues. The results will also cover each expert’s respective haptic applicationdesign scoring patterns on the technical, experience design, and design method score distri-butions used to address Focal Point 1.62How to Read Focal Point 4 VisualizationsCommunity Keyword Similarity Table Visualization1. A table of similarity values (proportion of shared keyword classes between two venues)will be shown. This can be seen in Figure 5.11. The keywords were found in Stage 2of our methodology.2. Higher similarity values indicate higher number of shared keywords between two venues.Higher saturation will also be used in these cases.3. Different colours are also used to denote the different communities.Haptic Expert Rating Pattern Visualizations1. Each expert’s rating patterns are shown side by side.2. Paper count (the number of papers that fell under a specific rating category) arereported as the y-axis of the figures described in Fig 5.143. Different colours indicate different communities.4. Trendlines indicate the general pattern (shape).5. Score categories represent differences in the levels of detail provided by a community,where higher score categories indicate higher levels of details being provided.6. We provide initial impressions of the data, without jumping into full discussion.7. Percentages Are Not Reported. These results report paper occurrences as hardcounts, not as percentages. We wanted to emphasize any possible disparities/gapsbetween the heights of different score categories.5.3.1 Focal Point 4 - Similarity of Words to Describe Haptic Ap-plication DesignAs a reminder, keyword classes were author simplified representations of the original longerkeywords that were thought to indicate an aspect of haptic application design. We chosethe keywords classes specifically for each community, based on our interpretations of whichwords implied haptic application design in the context of that community. Please refer toStage 2 Step 2B of Chapter 4 Methodology for the thought process behind the creation ofkeyword classes.63Figure 5.11 details our calculated similarity values for every pair of venues. These sim-ilarity values can be considered as the proportion of a venue pair’s shared keyword classesover the total unique keyword classes of the venue pair. Please refer back to Fig 4.11 fordetails as to how keyword similarity values were calculated.Figure 5.12 displays the shared keyword classes found between every pairing of com-munity venues. This will be useful in seeing the specific types of keyword classes that arecommon/different between the HCI and Haptics communities. In Fig 5.12, the values inthe cells represent the number of papers of a given venue-year that contained the keywordclass. The keyword classes listed for each comparison box were identified as common tothe two listed venue-years. For example, although the left-hand boxes show CHI 2014, thekeyword class list differs because CHI 2014 shared different keywords with CHI 2018 thanwith UIST 2018. Also seen in Figure 5.12, the list of shared keyword classes are shorter forcross-community comparison boxes than for comparisons within a community.Figure 5.11: The results of our keyword similarity analysis. Orange represents HCI compar-isons. Blue represents Haptics comparisons. Green represents cross-community comparisons.Higher saturation indicates higher similarity. Please refer to Fig 4.11 for similarity value cal-culation details.64Figure 5.12: Keyword class appearances for all analyzed venue pairs. Positioning and colour scheme intentionally resembles Fig5.11 for consistent visual language. However, the saturation scheme is not implemented in this figure.65HCI Venue Keyword Similarity Analysis - Based on the similarity value tablein Fig 5.11, the HCI community is more similar to each other than with Haptics venuecombinations. The HCI venue pairings showed an average similarity value of 0.26.Haptics Venue Keyword Similarity Analysis - For Haptics community venuecombinations, the resulting average similarity value was 0.10. This lower value maybe at-tributable to less consistent keyword classes being used by the Haptics venues.Cross-Community Venue Keyword Similarity Analysis - Typically, HCI com-munity and Haptics community combinations had an average similarity value of 0.01.Overall Keyword Class Similarity Analysis ImpressionsAfter the main author manually derived and analyzed the similarity of keyword classesaccording the process described in Section 4.64, the resulting average keyword class similarityvalues suggested low commonality in the language shared between the HCI and Hapticscommunities.Possible Explanation for Low Cross-Community Keyword Class SimilarityAs indicated by the shared keyword classes represented by Fig 5.12, we found that our HCI-venue keyword classes often referred to technology-based words, while Haptics-venue keywordclasses referred to activity-based words. For example, the HCI-venue keyword classes usedtechnology-based words such as “actuated” or “tactile”. The Haptics-venue keyword classesused activity-based words such as “surgical” or “training”. Thus, the nature of the typesof keyword classes appeared to be different for each community which may explain the lowcross-community similarity value.The low cross-community keyword class similarity value could also be attributed to thediverse nature of the Haptic-venue’s activity-based keyword classes. When looking at the key-word classes shared between ToH 2014/2018 and HS 2018 in Fig 5.12, certain ToH activity-based keyword classes such as “driving” or “game” did not appear in HS 2018. This mayhave also led to the Haptics community’s lower within community average keyword simi-larity value. Compared with the technology-based keyword classes used in the HCI venues,many keyword classes such as “actuated”, “haptic”, and “tactile” occurred more consistentlythroughout all HCI venues.5.3.2 Focal Point 4 - Common Words Used to Describe HapticApplication DesignFigure 5.13 shows the most popular shared keyword classes within the HCI, Haptics, andacross both communities.66Figure 5.13: Popular keyword classes among the communities. Keyword classes are generalterms derived from original keyword data via analysis.For the HCI community, we found “haptic” to be the most frequently occuring keywordclass for the haptic application papers that our overall process identified.For the Haptics community, our process identified “training” as the most frequentlyoccurring keyword class to indicate haptic application design papers.Across communities, “tactile” was the most common keyword class that was identifiedby our process in connecting the communities on describing haptic application design.5.3.3 Focal Point 4 – Haptic Expert Rating PatternsAs described in Chapter 4 (Methodology), each haptic expert was tasked to independentlyrate haptic application design areas for 60 potential haptic application design papers. Figure5.14 compares each expert’s respective rating pattern in order to draw ideas of whether thecommunities viewed haptic application design practices similarly or differently.67Figure 5.14: Each expert’s rating pattern for each haptic application design area.Technical Design Rating Comparison - Among technical design ratings, the HCIexperts (1 and 2) appeared to rate technical design details favourably. This was indicated by68the similiar trendline shapes for 1 and 2 in Figure 5.14. But the Haptics expert (3) had morevariance in their ratings, possibly due to their technical background causing a discerning eyefor technical details. This was indicated by Expert 3’s multimodal trendline shape in Figure5.14.Experience Design Rating Comparison - Among experience design ratings, theHCI experts seemed to have opposite views. Expert 1 rated experience design details ofpapers to be generally lower, indicated by the left leaning score distribution caused by higheroccurrences Score 0 and 1 ratings. Meanwhile, Expert 2 and Expert 3 shared similar ratingpatterns, where they both viewed papers to generally hold good experience design details.This was indicated by the right leaning distributions of Experts 2 and 3, caused by higheroccurrences of Score 2 and 3 ratings.Design Method Rating Comparison - Among design method ratings, a similarpattern of disparate HCI expert views was seen again. Expert 1 rated design method detailsto be quite lower than Expert 2, as indicated by Expert 1’s left leaning distribution, causedby higher occurrences of Score 1 ratings. Expert 2 seemed to show a more even distributionof detail ratings, as indicated by a “steady” right leaning distribution shape, caused by highoccurrences of Score 1, 2, and 3 ratings. However, Expert 3 viewed most papers to hold agood level of design method details, indicated by the spike in the Score 2 rating category.Inter-rater Reliability - Krippendorff’s alpha was used to assess the inter-rater reli-ability among our experts on the 3 areas of haptic application design [48]. For each areaof haptic application design, each expert’s ratings of the 60 candidate haptic applicationdesign papers were used to calculate Krippendorff’s alpha. In other words, 180 data points(60 ratings of a given haptic application design area for 3 experts) were used to calculate 3different alpha values.For each area of haptic application design, Krippendorff’s alpha was as follows. Technicaldesign (0.208), Experience Design (0.352), Design Methods (0.397). These alpha values canbe interpreted as ”fair agreement” among our experts in their views on haptic applicationdesign details [49]. However, it has been recommended that an alpha value of 0.667 to 0.8would be more ideal values for inter-rater agreement [50].Overall Impressions - Based on this limited sample, there were hints that our expertswere different in their interpretations of haptic application design.69Chapter 6DiscussionThis chapter will discuss our results regarding our focal points and associated research ques-tions.We will discuss the nature of differences/similarities on several aspects of haptic applica-tion design such as the types of user applications being made, the hardware used, technicaland user experience details being given, and the description/perspectives from the commu-nities.We will also preface by outlining the limitations that should be taken into considera-tion when interpreting the discussion to follow. We will suggest future work to extend thefoundations and address the limitations of our work.6.1 Limitations of Our WorkThere are several limitations of our work that should be taken into consideration.We were working under a number of constraints as we prototyped a new methodology.We have a number of observations for ways our process could be improved. Having moreresources (eg. bigger author team) would have also been helpful.Our limitations and its effect on the validity of our findings are as follows:1. Limited Number of Venues - We used a scoping review methodology to system-atically identify a review set of papers from a large body of literature. In this firstexploratory pass, we applied this methodology to a limited set of venues so that theresulting review set would fit within our limited analysis resources.In general, we considered our scoping review methodology a success due to our numer-ous filtering and paper inclusion checking steps creating a manageable set of papers70for analysis. However, our final set of analyzed papers was too small to support gen-eralizable interpretations (e.g. inferential statistics). This limited the degree to whichwe could form insights into how design practices varied among the communities. Weemphasize that this was a factor of the size of our sample, and not the methodologyof identifying appropriate samples for review.2. Main Author Team Bias - Many mechanics of our scoping review relied on humandriven decisions regarding search terms (keywords) and the generation of qualitativecodes to characterize general community design practices.There was a risk of bias at play, where it was possible that another set of researcherswill have come to different interpretations of which papers should be analyzed. Thus,data such as keywords or qualitative codes would have been different as a result.Specifically, in Stage 2 the research team manually selected keywords as an inclusioncriteria by which to extract relevant papers from each venue, resulting in the 82 papersthat left Stage 2. This manual selection stage had a subjective aspect to it, and asa result, it is possible that we chose papers that were based more on our personalinterpretations of haptic application design concepts.We attempted to mitigate this bias through many team discussions that attemptedto look at each mechanic of our methodology and how different types of communitymembers may perceive them as representative of a community’s perspective of hapticapplication design. However, this conversation can be improved by directly consultingmore haptic experts from each community.3. Uneven and Limited Expert Representation - We only had one expert represent-ing the Haptics community. This was due to challenges in recruiting another Hapticsexpert within the timeframe of our study. Additionally, our experts had variance intheir years of academic experience in researching and publishing haptic applicationdesign work.Given all of this, the final sample of papers used for analysis should be taken withmore precaution in its characterizations of haptic application design practices. Futureiterations of our work could involve a larger, diverse, and more experienced set ofresearchers to ensure a more equal/experienced set of eyes on viewing haptic applica-tion design literature. Having more experts could also help clarify whether personalbackground and bias affects how haptic application design details are interpreted frompublications.71Given these notable limitations, the reader should take the following discussion as alludingrather than confirming any differences/similarities of community haptic application designpractices. More work is required to generalize our findings of community haptic applicationdesign practices. We will discuss how to correct the limitations of our work in the FutureWork section below.6.2 Focal Point 1 (Design Energy Distribution)Our results suggested that when viewing design energy from a myriad of perspectives (com-munity level, year level, venue size level), each community seemed to prioritize similaramounts of design energy into the areas of haptic application design with a few exceptions.On a community level, both communities appeared to provide similar levels of details inall areas of haptic application design.On a year level, 2014 also suggested that both communities appeared to provide similarlevels of details in all areas of haptic application design. By contrast in 2018, a divergenceseemed to appear where the HCI community exhibited a growth in details in all areas ofhaptic application design.From a venue size level perspective, among our bigger sized sample venues, it was sug-gested that the HCI community prioritized experience design details. The smaller sizedvenues suggested some differences in details being provided. The HCI community providedslightly more details in all haptic application design areas.When taken as a whole, it is suggested by our results that the HCI community seemsto prioritize slightly more experience design details, while the Haptics community seemsconsistent in providing good levels of design details in all areas of haptic application design.Furthermore, this maybe a recent pattern, given that in 2014 both communities seemedconsistent in their level of haptic application design details among the included papers.However, this pattern may simply be due to the way in which we presented the papersto our experts. 2014 papers were listed first, and while we did not stipulate a certain orderof papers to be checked, it maybe the case that the experts used the 2014 papers to create abaseline of their rating patterns. Thus, the year level differences should be taken cautiously.Due to our limitation of uneven group sizes in all of the comparisons, all of our designenergy analysis should be taken cautiously. In particular, certain community patterns ondesign energies can be inflated by the differences in group size.726.3 Focal Point 2 (Purpose of Haptics in User Experi-ences)Based the main author’s analysis, it was suggested that each community seemed to emphasizedifferent types of user experiences.The HCI community seemed to emphasize social communication, visual impairment sup-port, and professional physical support application experiences. The Haptics communityseemed to emphasize experiences for medical professionals and physical rehabilitation. Bothcommunities shared driving support experiences.Furthermore, it appears that the purported reasons to design with haptic technologycan be corroborated with these user experiences. The HCI community seemed to use haptictechnology as an information display to a greater extent than that of the Haptics community.Information displays can be a possible way in which social communication could thriveunder. The Haptics community seemed to use haptic technology to support improvementsin control to a bigger extent than that of the HCI community. The dexterity traininginvolved in medical professional and physical rehabilitation seems to be a natural fit forcontrol improvements.However, these interpretations should be taken cautiously. By virtue of our method-ology, our final sample size of papers was greatly reduced from a much larger space ofliterature. Thus, it is possible that different types of experiences and associated purposesexist. Ensuring a larger final sample of papers could be one way to discover whether otherexperiences/purposes exist.6.4 Focal Point 3 (Design Methods Used to CreateHaptic Applications)Our results suggested that each community provided different levels of focus for differentdesign methods.Within technical design, the HCI community seemed to focus on providing general systemimplementation details. In addition to general system implementation details, the Hapticscommunity appeared to give more specific technical design details in the form of algorithmsand stimuli parameters. Interestingly, the HCI community appeared to exhibit unique in-stances of design sketching system behaviour (eg. prototyping different types of hardwareor stimuli).In terms of describing their design methods, the HCI community appeared to provide73use-cases and possible design spaces to ratify the purpose of a haptic application to a muchlarger extent than the Haptics community. The Haptics community appeared report onthese details to a lower extent, perhaps due to a technological contribution being obviouslyapplicable for a certain class of user applications. It was also found that both communitiesdid not always report design method details. Occasionally, the HCI community also seemedto exhibit a few unique instances of brainstorming, rapid prototyping, co-design and lowfidelity prototyping.Each community also appeared to have distinct preferences for certain hardware. TheHCI community seemed to prefer tactile display technologies. The Haptics communityseemed to prefer force feedback technologies. Interestingly, both communities were alsointerested in vibrotactile display technologies. The HCI community appeared to have a fewunique instances of midair and dissipative technologies.These preferences for certain hardware could be tied to each community’s preferred userapplications. Medical professional, physical rehabilitation and driving applications that re-quire good control is more amenable towards force feedback technology, which can providehigher fidelity haptic effects. Meanwhile, information displays can be created using vibro-tactile or tactile display technologies, which is amenable to general user application channelssuch as smartphones.It is important to be cautious of these interpretations. It maybe the case that ourmethodology’s sampling methods resulted in these particular patterns to be exhibited whenother patterns were possible. Thus, it is important to examine a larger sample of papers inthe future.6.5 Focal Point 4 (Understanding Community Perspec-tives on Haptic Application Design)Each community appeared to use different types of keyword classes to describe haptic appli-cation design.This author team developed keyword classes out of the keywords that authors of theselected papers used to describe their own papers. Thus, there is some potential for subjec-tivity in the creation of these keyword classes. However, we see an opportunity for a largerconversation of what keyword classes should best represent the diverse perspectives of thesecommunities. But, keeping this in mind, we found different distributions of our keywordclasses in the haptic application design papers from each community.The HCI community keyword classes appeared to involve haptic technology, and not74typically words related to interaction or user experiences. This could be due to the implicitinteraction aspect of the HCI community.The Haptics community keyword classes appeared to evoke activity-based keywords,rather than words regarding haptic technology. This could be due to the implicit tech-nological aspect of the Haptics community.It is possible that differences in the community backgrounds, described in Chapter 2(Background), could partially explain the differences we saw in how haptic application designwas described.It is difficult to say whether the haptic experts’ rating data indicated individual patternsand biases in their ratings, given the small number of raters employed. There is no doubtthat their ratings differed, but we cannot say how these might link to factors like disciplinarybackground and specifics of their own design experience versus the personal tendencies wemight see in any kind of rating tasks (eg. a tendency to rate more harshly or generously)without larger rater numbers.6.6 Overall Impressions - Are the Communities Differ-ent or Similar to Each Other?Our methodology and results seemed to suggest that the HCI and Haptics communitiespracticed haptic application design more differently rather than similarly.In terms of the design energy, possible differences seemed to exist in terms of the HCIcommunity investing more detail in user experience design details. In terms of the keywordsused to report haptic application design work, the communities did not appear to sharemuch language with each other. However, this maybe due to a bias in main author keywordselection.The most notable differences appeared to be the different types of user applications beingdesigned, the purposes of why haptic technology was used in a design, and the technologi-cal preferences of the communities. While often both communities exhibited many sharedexamples in applications and technology, the differences appeared in the extent of papersreporting such details. The potential source of these differences could be due to the respec-tive backgrounds of each community informing what type of haptic application details areimportant.Some similarities were noticed. In terms of design methods being reported, each com-munity seemed to place more value on providing technical oriented details, with the HCIcommunity providing general system details while the Haptics community provided more75specific technical implementation details. Vibrotactile technology appeared to be a populartechnology that was used in almost equal degrees by both communities.6.7 Future Work - Suggestions for Improving Our StudyWe recommend the following next steps to address the limitations of our study in two keyways. The first way being to improve the methodology of the current work, and the secondway suggesting how to use this work for different types of studies.Improving Methodology of Current Work1. Expand the Sample Size and Groups Being Studied - The confidence of ourresults can increase by expanding the number of papers and community groups. Wedescribe a possible approach using Figure 8.1 in the Appendix section.2. Involve More Community Experts - Our current study only involved 3 experts.We also did not have equal representation of the communities among our experts (2HCI, 1 Haptics). Thus, having equal numbers of experts representing the communitiescan help gain more fair views on haptic application design.Suggestions for Different Studies1. Investigate How Much ”Cross-Fertilization” is Happening - Our study pro-vided a very high level look on how communities practiced haptic application design.It would be interesting to investigate how researchers with a mixed HCI and Hapticscommunity background approaches haptic application design projects. This investiga-tion could be started by examining whether authors of the venues used in our studyhave published in other community venues.2. Open the Discussion on the Differences/Similarities of Haptic ResearchCommunities - We hope that this research shed some light to the nature of designdifferences found between the communities. It would be interesting to see the differentcommunity members talk to each other about our findings in order to obtain ideas andopportunities for haptic application design knowledge sharing.76Chapter 7ConclusionsOur research goal was to discover the different/similar community practices used to designhaptic applications. Specifically, we wished to gain insights into design practices regardingthe technical design, experience design, and design methods used to create haptic applica-tions. We contributed a novel scoping literature review methodology to carefully find andcompare haptic application design papers representative of the communities.Our results contributed an initial characterization of haptic application design practicescommon among the communities. Our results indicated many areas of differences ratherthan similarities.As mentioned in our introduction, our results have interesting implications for the com-munities in terms of opportunities for knowledge sharing. The nature of the differences wasof an complementary nature, that could allow for one community’s knowledge to “fill in thegaps” of another community. For example, the HCI community could leverage the excellenttechnical details of the Haptics community to improve the saliency of haptic information dis-plays. The Haptics community could leverage the ideation and use-case defining approachesof the HCI community to create experiences accessible to a broader audience. Our results,while preliminary, suggest that there is potential for each community to leverage each other’sstrengths in order to improve or create new types of haptic application experiences.However, these implications will need to be investigated further. The expansion of papersand different community venues can help verify the generalizability of our results. Investi-gating mixed community background researcher’s approaches to haptic application designcan detail the benefits/specifics of design knowledge cross-fertilization.Despite the limitations and scale of this study, we believe the future is bright for theHCI and Haptics communities. We are excited by the possibilities of new meaningful hapticapplication experiences for when the communities are ready to collaborate with each other.The seeds of knowledge are just waiting to be shared and grown.77Bibliography[1] “AI Applications,” accessed Feb 2020. [Online]. Available: https://towardsdatascience.com/7-impacts-of-the-artificial-intelligence-technology-5a4663397961[2] J. Preece, Y. Rogers, and H. Sharp, Interaction design: Beyond human-computer inter-action. John Wiley & Sons, 2015.[3] “World Haptics,” accessed Nov 2019. [Online]. Available: http://www.worldhaptics.org/content/haptic-conferences[4] “Design Thinking Process,” accessed Aug 2019. [Online]. Available: https://www.interaction-design.org/literature/article/5-stages-in-the-design-thinking-process[5] V. Braun and V. Clarke, “Using thematic analysis in psychology,” Qualitative Researchin Psychology, vol. 3, no. 2, pp. 77–101, 2006.[6] “ASME,” accessed Nov 2019. [Online]. Available: https://www.mitpressjournals.org/loi/pvar[7] “Haptics Conference History,” accessed March 2019. [Online]. Available: http://about.hapticssymposium.org/[8] “Other Haptic Journals/Conferences,” accessed Nov 2019. [Online]. Available:https://ieeexplore.ieee.org/xpl/aboutJournal.jsp?punumber=10[9] “CS 543 Haptic Design Course,” accessed Oct 2019. [Online]. Available: https://www.cs.ubc.ca/∼cs543/current-term/[10] “EuroHaptics,” accessed Jan 2020. [Online]. Available: https://eurohaptics.org/conferences/[11] “World Haptics Asia,” accessed Jan 2020. [Online]. Available: http://www.worldhaptics2019.org/78[12] K. E. MacLean, O. S. Schneider, and H. Seifi, Multisensory Haptic Interactions:Understanding the Sense and Designing for It. Association for ComputingMachinery and Morgan Claypool, 2017, p. 97–142. [Online]. Available: https://doi.org/10.1145/3015783.3015788[13] J. H. Bae, C. J. Ploch, M. A. Lin, B. L. Daniel, and M. R. Cutkosky, “Display of needletip contact forces for steering guidance,” in HAPTICS Haptics Symp. IEEE, 2016, pp.527–532.[14] V. Girbes, L. Armesto, J. Dols, and J. Tornero, “Haptic feedback to assist bus driversfor pedestrian safety at low speed,” IEEE Trans. on Haptics, vol. 9, no. 3, pp. 345–357,2016.[15] Y. Wang and K. J. Kuchenbecker, “Halo: Haptic alerts for low-hanging obstacles inwhite cane navigation,” in HAPTICS Haptics Symp. IEEE, 2012, pp. 527–532.[16] M. Gabardi, M. Solazzi, D. Leonardis, and A. Frisoli, “A new wearable fingertip hapticinterface for the rendering of virtual shapes and surface features,” in HAPTICS HapticsSymp. IEEE, 2016, pp. 140–146.[17] J. Lee, B. Han, and S. Choi, “Motion effects synthesis for 4d films,” IEEE Trans. onHaptics, vol. 22, no. 10, pp. 2300–2314, 2016.[18] J. Lee, J. Ryu, and S. Choi, “Graphical authoring tools for vibrotactile patterns,” inWHC World Haptics Third Joint EuroHaptics Conf. and Symp. on Haptic Interfacesfor Virtual Environment and Teleoperator Systems. IEEE, 2009, pp. 388–389.[19] F. Danieau, J. Fleureau, P. Guillotel, N. Mollet, M. Christie, and A. Le´cuyer, “Towardhaptic cinematography: enhancing movie experiences with camera-based haptic effects,”IEEE MultiMedia, vol. 21, no. 2, pp. 11–21, 2014.[20] M. J. Enriquez and K. E. MacLean, “The hapticon editor: a tool in support of hapticcommunication research,” in 11th Symp. on Haptic Interfaces for Virtual Environmentand Teleoperator Systems. IEEE, 2003, pp. 356–362.[21] O. S. Schneider and K. E. MacLean, “Studying design process and example use withmacaron, a web-based vibrotactile effect editor,” in HAPTICS Haptics Symp. IEEE,2016, pp. 52–58.[22] H. Seifi, C. Anthonypillai, and K. E. MacLean, “End-user customization of affectivetactile messages: A qualitative examination of tool parameters,” in HAPTICS HapticsSymp. IEEE, 2014, pp. 251–256.79[23] H. Seifi, K. Zhang, and K. E. MacLean, “Vibviz: Organizing, visualizing and navigatingvibration libraries,” in WHC World Haptics Conf. IEEE, 2015, pp. 254–259.[24] R. Ueoka, M. Yamaguchi, and Y. Sato, “Interactive cheek haptic display with air vortexrings for stress modification,” in CHI Proc. of the Conf. Extended Abstracts on HumanFactors in Computing Systems. ACM, 2016, pp. 1766–1771.[25] J. Allen, L. Cang, M. Phan-Ba, A. Strang, and K. MacLean, “Introducing the cuddlebot:A robot that responds to touch gestures,” in Proc. of the Tenth Annual InternationalConf. on Human-Robot Interaction Extended Abstracts. ACM/IEEE, 2015, pp. 295–295.[26] D. Tam, K. E. MacLean, J. McGrenere, and K. J. Kuchenbecker, “The design and fieldobservation of a haptic notification system for timing awareness during oral presenta-tions,” in SIGCHI Proc. of the Conf. on Human Factors in Computing Systems. ACM,2013, pp. 1689–1698.[27] M. K. Pan, J. McGrenere, E. A. Croft, and K. E. MacLean, “Exploring the role ofhaptic feedback in enabling implicit hci-based bookmarking,” IEEE Trans. on Haptics,vol. 7, no. 1, pp. 24–36, 2014.[28] K. Minamizawa, Y. Kakehi, M. Nakatani, S. Mihara, and S. Tachi, “Techtile toolkit:a prototyping tool for design and education of haptic media,” in Proc. of the VirtualReality International Conf. ACM, 2012, p. 26.[29] H. Seifi, F. Fazlollahi, M. Oppermann, J. A. Sastrillo, J. Ip, A. Agrawal, G. Park, K. J.Kuchenbecker, and K. E. MacLean, “Haptipedia: Accelerating haptic device discoveryto support interaction & engineering design,” in CHI Proc. of the Conf. on HumanFactors in Computing Systems. ACM, 2019, p. 558.[30] A. Hamam, M. Eid, A. El Saddik, and N. D. Georganas, “A quality of experience modelfor haptic user interfaces,” in Proc. of the Ambi-Sys Workshop on Haptic User Interfacesin Ambient Media Systems. ICST (Institute for Computer Sciences, Social-Informaticsand Telecommunications Engineering), 2008, p. 1.[31] A. Pusch and A. Le´cuyer, “Pseudo-haptics: from the theoretical foundations to practi-cal system design guidelines,” in Proc. of the 13th International Conf. on MultimodalInterfaces. ACM, 2011, pp. 57–64.80[32] D. Lo¨ffler, R. Tscharn, and J. Hurtienne, “Multimodal effects of color and haptics onintuitive interaction with tangible user interfaces,” in Proc. of the Twelfth InternationalConf. on Tangible, Embedded, and Embodied Interaction. ACM, 2018, pp. 647–655.[33] S. M. Petermeijer, D. A. Abbink, M. Mulder, and J. C. de Winter, “The effect of hapticsupport systems on driver performance: A literature survey,” IEEE Trans. on Haptics,vol. 8, no. 4, pp. 467–479, 2015.[34] G. Ravali and M. Manivannan, “Haptic feedback in needle insertion modeling and sim-ulation,” IEEE Reviews in Biomedical Engineering, vol. 10, pp. 63–77, 2017.[35] J. A. Griffin, W. Zhu, and C. S. Nam, “The role of haptic feedback in robotic-assistedretinal microsurgery systems: a systematic review,” IEEE Trans. on Haptics, vol. 10,no. 1, pp. 94–105, 2016.[36] K. S. Hale and K. M. Stanney, “Deriving haptic design guidelines from human physi-ological, psychophysical, and neurological foundations,” IEEE Computer Graphics andApplications, vol. 10, no. 1, pp. 94–105, 2016.[37] C. Sjostrom, “Designing haptic computer interfaces for blind people,” in Proc. of theSixth International Symp. on Signal Processing and its Applications. IEEE, 2001, pp.68–71.[38] J. Song, J. H. Lim, and M. H. Yun, “Finding the latent semantics of haptic interactionresearch: A systematic literature review of haptic interaction using content analysisand network analysis,” Human Factors and Ergonomics in Manufacturing & ServiceIndustries, vol. 26, no. 5, pp. 577–594, 2016.[39] S. Coemans and K. Hannes, “Researchers under the spell of the arts: Two decades ofusing arts-based methods in community-based inquiry with vulnerable populations,”Educational Research Review, vol. 22, pp. 34–49, 2017.[40] “CHI Committees,” accessed Nov 2019. [Online]. Available: http://chi2019.acm.org/authors/papers/selecting-a-subcommittee/[41] “Literature Review Guide,” accessed Nov 2019. [Online]. Available: http://guides.library.ubc.ca/litreviews[42] J. Jesson, L. Matheson, and F. M. Lacey, Doing Your Literature Review: Traditionaland Systematic Techniques. Sage, 2011.81[43] Z. Munn, M. D. Peters, C. Stern, C. Tufanaru, A. McArthur, and E. Aromataris,“Systematic review or scoping review? Guidance for authors when choosing between asystematic or scoping review approach,” BMC Medical Research Methodology, vol. 18,no. 1, p. 143, 2018.[44] D. Moher, A. Liberati, J. Tetzlaff, and D. G. Altman, “Preferred reporting items for sys-tematic reviews and meta-analyses: the prisma statement,” Annals of Internal Medicine,vol. 151, no. 4, pp. 264–269, 2009.[45] A. Kitson, M. Prpa, and B. E. Riecke, “Immersive interactive technologies for positivechange: a scoping review and design considerations,” Frontiers in Psychology, vol. 9, p.1354, 2018.[46] H. Arksey and L. O’Malley, “Scoping studies: towards a methodological framework,”International Journal of Social Research Methodology, vol. 8, no. 1, pp. 19–32, 2005.[47] A. C. Tricco, E. Lillie, W. Zarin, K. K. O’Brien, H. Colquhoun, D. Levac, D. Moher,M. D. Peters, T. Horsley, L. Weeks, S. Hempel, E. A. Akl, C. Chang, J. McGowan,L. Stewart, L. Hartling, A. Aldcroft, M. G. Wilson, C. Garritty, S. Lewin, C. M. Godfrey,M. T. Macdonald, E. V. Langlois, K. Soares-Weiser, J. Moriarty, T. Clifford, Tunc¸alp,and S. E. Straus, “PRISMA Extension for Scoping Reviews (PRISMA-ScR): Checklistand Explanation,” Annals of Internal Medicine, vol. 169, no. 7, pp. 467–473, 2018.[48] “Krippendorff’s Alpha Background,” accessed Feb 2020. [Online]. Available:https://repository.upenn.edu/cgi/viewcontent.cgi?article=1043&context=asc papers[49] “Interpreting Reliability Results,” accessed Feb 2020. [Online]. Available: http://homepages.inf.ed.ac.uk/jeanc/maptask-coding-html/node23.html[50] “Krippendorff’s Alpha,” accessed Feb 2020. [Online]. Available: https://www.statisticshowto.datasciencecentral.com/krippendorffs-alpha/[51] P. Lopes, D. Yu¨ksel, F. Guimbretie`re, and P. Baudisch, “Muscle-plotter: An interactivesystem based on electrical muscle stimulation that produces spatial output,” in UISTProc. of the 29th Annual Symp. on User Interface Software and Technology. ACM,2016, pp. 207–217.[52] A. Tanaka and A. Parkinson, “Haptic wave: A cross-modal interface for visually im-paired audio producers,” in CHI Proc. of the Conf. on Human Factors in ComputingSystems. ACM, 2016, pp. 2150–2161.82[53] O. Schneider, K. MacLean, C. Swindells, and K. Booth, “Haptic experience design:What hapticians do and where they need help,” International Journal of Human-Computer Studies, vol. 107, pp. 5–21, 2017.83Chapter 8Appendix8.1 Future Work - Systematically Finding AppropriateVenuesIt was described in Chapter 4 (Methodology) that the main author team’s personal knowledgeserved as the primary basis for choosing the venues of our study.After completing this study, we believe the following approach could be used in futurework to methodically pick new research venues.Figure 8.1 details a possible approach in utilizing “yields”, an estimation of a venue’snumber of haptic application design papers.How to Calculate Yield - Researchers can first find the total number of papers frompossible haptic application design venues. Next, an estimation of the number of haptic ap-plication design papers could be made using personal knowledge and expert advice regardingthe possible venues. Lastly, the resulting yield (proportion of haptic application design pa-pers over a venue’s total paper set) can then be used as a measure to decide whether a venueis worthwhile for analysis.84Figure 8.1: A possible approach to find new venues for analysis. The actual number of hapticapplication design papers found are used for illustration.8.2 Papers Used to Generate Focal PointsThe following example papers were used to generate our focal points regarding the technical,user experience and design methods of haptic applications practiced in both communities.These papers were chosen from notable haptic research venues based on our personal knowl-edge. However, the year of the papers were not the same as the final years chosen in ourmethodology. The year span used in our methodology was intentionally different.85Venue/Year Focal Point Generating PublicationCHI 2016S. Kratz and A. Dunnigan,”Thermotouch: Design of a high dynamic temperature range thermal haptic display,”in CHI Proc. Conf. Extended Abstracts on Human Factors in Computing Systems. ACM, 2016, pp. 1577-1582.CHI 2016S. Khurelbaatar et al.”Tactile presentation to the back of a smartphone with simultaneous screen operation,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2016, pp. 3717-3721.CHI 2016A. Tanaka and A. Parkinson,”Haptic wave: A cross-modal interface for visually impaired audio producers,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2016, pp. 2150-2161.CHI 2016M. Ernst and A. Girouard,”Exploring haptics for learning bend gestures for the blind,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2016, pp. 2097-2104.CHI 2016J. Nishida and K. Suzuki,”Biosync: Synchronous kinesthetic experience among people,”in CHI Proc. Conf. Extended Abstracts on Human Factors in Computing Systems. ACM, 2016, pp. 3742-3745.CHI 2016R. Ueoka, M. Yamaguchi, and Y. Sato,”Interactive Cheek Haptic Display with Air Vortex Rings for Stress Modification,”in CHI Proc. Conf. Extended Abstracts on Human Factors in Computing Systems. ACM, 2016, pp. 1766-1771.CHI 2016A. Gupta et al.”Direct manipulation in tactile displays,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2016, pp. 3683-3693.UIST 2016J. Ou et al.”aeroMorph-heat-sealing inflatable shape-change materials for interaction design,”in UIST Proc. 29th Annual Symp. on User Interface Software and Technology. ACM, 2016, pp. 121-132.UIST 2016Y. Cho et al.”RealPen: Providing realism in handwriting tasks on touch surfaces using auditory-tactile feedback,”in UIST Proc. 29th Annual Symp.on User Interface Software and Technology. ACM, 2016, pp. 195-205.UIST 2016J. Gugenheimer et al.”Gyrovr: Simulating inertia in virtual reality using head worn flywheels,”in UIST Proc. 29th Annual Symp. on User Interface Software and Technology. ACM, 2016, pp. 227-232.UIST 2016P. Lopes et al.”Muscle-plotter: An interactive system based on electrical muscle stimulation that produces spatial output,”in UIST Proc. 29th Annual Symp. on User Interface Software and Technology. ACM, 2016, pp. 207-217.UIST 2016A. Dementyev et al.”Rovables: Miniature on-body robots as mobile wearables,”in UIST Proc. 29th Annual Symp. on User Interface Software and Technology. ACM, 2016, pp. 111-120.Table 8.1: HCI community example papers used to inspire our focal points.86Venue/Year Focal Point Generating PublicationHS 2016D. Spelmezan et al.”SkinHaptics: Ultrasound focused in the hand creates tactile sensations,”in HAPTICS Haptics Symp. IEEE, 2016, pp. 98-105.HS 2016M. Gabardi et al.”A new wearable fingertip haptic interface for the rendering of virtual shapes and surface features,”in HAPTICS Haptics Symp. IEEE, 2016, pp. 140-146.HS 2016A. Russomanno et al.”Modeling latching fluidic circuits to determine clocking limits for a refreshable braille display,”in HAPTICS Haptics Symp. IEEE, 2016, pp. 179-184.HS 2016M. Price and F.C. Sup,”A robotic touchscreen totem for two-dimensional haptic force display,”in HAPTICS Haptics Symp. IEEE, 2016, pp. 72-77.HS 2016A.J. Spiers and A.M. Dollar,”Outdoor pedestrian navigation assistance with a shape-changing haptic interface and comparison with a vibrotactile device,”in HAPTICS Haptics Symp. IEEE, 2016, pp. 34-40.HS 2016J.H. Bae et al.”Display of needle tip contact forces for steering guidance,”in HAPTICS Haptics Symp. IEEE, 2016, pp. 332-337.ToH 2016F. Dimeas and N. Aspragathos,”Online stability in human-robot cooperation with admittance control,”IEEE Trans. on Haptics, vol.9, no.2, pp. 267-278, 2016.ToH 2016S. Yim, S. Jeon and S. Choi,”Data-driven haptic modeling and rendering of viscoelastic and frictional responses of deformable objects,”IEEE Trans. on Haptics, vol.9, no.4, pp. 548-559, 2016.ToH 2016P. Olsson et al.”Comparison of walking and traveling-wave piezoelectric motors as actuators in kinesthetic haptic devices,”IEEE Trans. on Haptics, vol.9, no.3, pp. 427-431, 2016.ToH 2016V. Girbes et al.”Haptic feedback to assist bus drivers for pedestrian safety at low speed,”IEEE Trans. on Haptics, vol.9, no.3, pp. 345-357, 2016.ToH 2016S. Okamoto, M. Wiertlewski and V. Hayward,”Anticipatory vibrotactile cueing facilitates grip force adjustment during perturbative loading,”IEEE Trans. on Haptics, vol.9, no.2, pp. 233-242, 2016.ToH 2016J. Lee, Y. Kim and G.J. Kim,”Rich pinch: Perception of object movement with tactile illusion,”IEEE Trans. on Haptics, vol.9, no.1, pp. 80-89, 2016.Table 8.2: Haptics community example papers used to inspire our focal points.8.3 Papers Used in Methodology and AnalysisOur scoping review involved many stages filtering out papers from a large research space.This was in order to find the most suitable papers appropriate for analysis. Please refer toChapter 4 (Methodology) for specific details.Stage 2 of our methodology established a possible set of 82 haptic application designpapers inferred from analyzing the keyword metadata of our venue’s papers. This camefrom the rejection of 1298 papers from the original number of papers provided by each ofour venues.While Stage 2 used a mixed automatic and human decision approach to filter out papers.The subsequent stages involved much more human interpretations and judgements thatwarrants examples that illustrates what sort of papers tended to be dropped from our study.87Stage 3 of our methodology involved the first author reading the titles/abstracts of Stage2’s 82 papers in order to make a judgement for further inclusion in the study. Table 8.3shows the 22 dropped papers that were considered not appropriate for study inclusion. Thesepapers involved the HCI community exclusively. After the 22 papers were dropped, 60 papersremained possible candidates for further analysis.88Venue/Year Excluded PublicationCHI 2014L.P. Cheng et al.”Haptic turk: a motion platform based on people,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 3463-3472.CHI 2018H. Kim et al.”Hapcube: A wearable tactile device to provide tangential and normal pseudo-force feedback on a fingertip,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-13.CHI 2018A. Niijima et al.”Controlling maximal voluntary contraction of the upper limb muscles by facial electrical stimulation,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-7.CHI 2018K. Kato et al.”Double-sided printed tactile display with electro stimuli and electrostatic forces and its assessment,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-12.CHI 2018I.Choi et al.”Claw: A multifunctional handheld haptic controller for grasping, touching, and triggering in virtual reality,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-13.CHI 2018A. Delazio et al.”Force jacket: Pneumatically-actuated jacket for embodied haptic experiences,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-12.CHI 2018S. Heo et al.”Thor’s hammer: An ungrounded force feedback device utilizing propeller-induced propulsive force,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-11.CHI 2018E. Strasnick et al.”Haptic links: Bimanual haptics for virtual reality using variable stiffness actuation, ”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-12.CHI 2018G. Park and S. Choi,”Tactile information transmission by 2d stationary phantom sensations,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-12.CHI 2018P. Abtahi and S. Follmer,”Visuo-haptic illusions for improving the perceived performance of shape displays,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-13.CHI 2018Y.A. Shim, J. Lee and G. Lee,”Exploring multimodal watch-back tactile display using wind and vibration,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-12.CHI 2018G. Wang et al.”Printed paper actuator: A low-cost reversible actuation and sensing method for shape changing interfaces,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-12.CHI 2018E. Whitmire et al.”Haptic revolver: Touch, shear, texture, and shape rendering on a reconfigurable virtual reality controller,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-12.CHI 2018D.R. Sahoo et al.”Tangible drops: a visio-tactile display using actuated liquid-metal droplets,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 1-14.UIST 2018A. Withana, D. Groeger and J. Steimle,”Tacttoo: A thin and feel-through tattoo for on-skin tactile output,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 365-378.UIST 2018J.Y. Lo et al.”RollingStone: Using Single Slip Taxel for Enhancing Active Finger Exploration with a Virtual Reality Controller,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 839-851.UIST 2018D.Y. Huang et al.”Orecchio: Extending body-language through actuated static and dynamic auricular postures,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 697-710.UIST 2018H.Y. Chang et al.”Facepush: Introducing normal force on face with head-mounted displays,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 927-935.UIST 2018T. Han et al.”HydRoring: Supporting mixed reality haptics using liquid flow,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 913-925.UIST 2018R Hinchet et al.”Dextres: Wearable haptic feedback for grasping in vr via a thin form-factor electrostatic brake,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 901-912.UIST 2018B. Son and J. Park,”Haptic feedback to the palm and fingers for improved tactile perception of large objects,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 757-763.UIST 2018G. Wang et al.”4DMesh: 4D printing morphing non-developable mesh surfaces,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 623-635.Table 8.3: The 22 excluded papers based on title/abstract checking by the first author.89Stage 3 also involved different haptic experts reading the 60 papers produced by ti-tle/abstract checking. Table 8.4 and Table 8.5 shows the 14 dropped papers across theHCI and Haptics communities that were not considered appropriate for study inclusion bythe experts. After experts checked the papers, 46 papers were considered suitable for finalanalysis.Venue/Year Excluded PublicationCHI 2014H. Nakanishi, K. Tanaka and Y. Wada,”Remote handshaking: touch enhances video-mediated social telepresence,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 2143-2152.CHI 2014J. Kangas et al.”Gaze gestures and haptic feedback in mobile devices,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 435-438.CHI 2014P. Dimitriadis and J. Alexander,”Evaluating the effectiveness of physical shape-change for in-pocket mobile device notifications,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 2589-2592.UIST 2018M. Ogata,”Magneto-Haptics: Embedding Magnetic Force Feedback for Physical Interactions,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 737-743.Table 8.4: The 4 HCI papers deemed not suitable for further analysis by our experts. Thisshows 4 out of 14 papers deemed not suitable for further analysis.90Venue/Year Excluded PublicationToH 2014C. Ho, R. Gray and C. Spence,”Reorienting driver attention with dynamic tactile cues,”IEEE Trans. on Haptics, vol.7, no.1, pp. 86-94, 2014.ToH 2014F. Mars, M. Deroo and J.M. Hoc,”Analysis of human-machine cooperation when driving with different degrees of haptic shared control,”IEEE Trans. on Haptics, vol.7, no.3, pp. 324-333, 2014.ToH 2014F. Ve´rite´, W. Bachta and G. Morel,”Closed loop kinesthetic feedback for postural control rehabilitation,”IEEE Trans. on Haptics, vol.7, no.2, pp. 150-160, 2014.ToH 2014L. Corenthy et al.”Haptically assisted connection procedure for the reconstruction of dendritic spines,”IEEE Trans. on Haptics, vol.7, no.4, pp. 486-498, 2014.ToH 2014S. Scheggi, F. Morbidi and D. Prattichizzo,”Human-robot formation control via visual and vibrotactile haptic feedback,”IEEE Trans. on Haptics, vol.7, no.4, pp. 499-511, 2014.ToH 2014D. Wang et al.”Haptic simulation of organ deformation and hybrid contacts in dental operations,”IEEE Trans. on Haptics, vol.7, no.1, pp. 48-60, 2014.HS 2018J. Anlauff, T. Kim, and J.R. Cooperstock,”Feel-a-bump: Haptic feedback for foot-based angular menu selection,”in HAPTICS Haptics Symp. IEEE, 2018, pp. 175-179.HS 2018S.M. Sketch, A.J. Bastian and A.M. Okamura,”Comparing proprioceptive acuity in the arm between joint space and task space,”in HAPTICS Haptics Symp. IEEE, 2018, pp. 125-132.HS 2018D.C. Egloff, M.M. Wanderley and I. Frissen,”Haptic display of melodic intervals for musical applications,”in HAPTICS Haptics Symp. IEEE, 2018, pp. 284-289.HS 2018H. Culbertson et al.”A social haptic device to create continuous lateral motion using sequential normal indentation,”in HAPTICS Haptics Symp. IEEE, 2018, pp. 32-39.Table 8.5: The 10 Haptics papers deemed not suitable for further analysis by our experts.This shows 10 out of 14 papers deemed not suitable for further analysis.Before analysis of the 46 expert checked papers was fully conducted, the first authorconducted a final paper suitability check. From this check, 8 papers were dropped. Table8.6 and Table 8.7 shows the 8 dropped papers across the HCI and Haptics communities thatwere not considered appropriate for final analysis by the first author.Venue/Year Excluded PublicationUIST 2018S.Y. Teng et al.”Pupop: Pop-up prop on palm for virtual reality,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 5-17.Table 8.6: The single HCI paper deemed not suitable for final analysis by the first author.This shows 1 out of 8 papers deemed not suitable for final analysis.91Venue/Year Excluded PublicationToH 2014S. Jeon and M. Harders,”Haptic tumor augmentation: Exploring multi-point interaction,”IEEE Trans. on Haptics, vol.7, no.4, pp. 477-485, 2014.ToH 2014F. Gonzalez, F. Gosselin and W. Bachta,”Analysis of hand contact areas and interaction capabilities during manipulation and exploration,”IEEE Trans. on Haptics, vol.7, no.4, pp. 415-429, 2014.ToH 2014C.K. Williams and H. Carnahan,”Motor learning perspectives on haptic training for the upper extremities,”IEEE Trans. on Haptics, vol.7, no.2, pp. 240-250, 2014.ToH 2014M. Azadi and L.A. Jones,”Evaluating vibrotactile dimensions for the design of tactons,”IEEE Trans. on Haptics, vol.7, no.1, pp. 14-23, 2014.ToH 2014N. Garcia-Hernandez et al.”How tactor size and density of normal indentation tactile displays affects grating discrimination tasks,”IEEE Trans. on Haptics, vol.7, no.3, pp. 356-366, 2014.ToH 2018M. Ogrinc et al.”Sensory integration of apparent motion speed and vibration magnitude,”IEEE Trans. on Haptics, vol.11, no.3, pp. 455-463, 2018.HS 2018J. Jiao et al.”Data-driven rendering of fabric textures on electrostatic tactile displays,”in HAPTICS Haptics Symp. IEEE, 2018, pp. 169-174.Table 8.7: The 7 Haptics papers deemed not suitable for final analysis by the first author.This shows 7 out of 8 papers deemed not suitable for final analysis.We used the following 38 papers in our final analysis described in Chapter 4 (Methodol-ogy). Table 8.8 shows the final HCI community papers. Table 8.9 shows the final Hapticscommunity papers.92Venue/Year Analyzed PublicationCHI 2014T. Hachisu and M. Fukumoto,”VacuumTouch: attractive force feedback interface for haptic interactive surface using air suction,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 411-420.CHI 2014E. Gronvall et al.”Causing commotion with a shape-changing bench: experiencing shape-changing interfaces in use,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 2559-2568.CHI 2014M. Prasad et al.”Haptimoto: Turn-by-turn haptic route guidance interface for motorcyclists,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 3597-3606.CHI 2014G. Wilson et al.”Perception of ultrasonic haptic feedback on the hand: localisation and apparent motion,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 1133-1142.CHI 2014J. Park et al.”Wrigglo: shape-changing peripheral for interpersonal mobile communication,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 3973-3976.CHI 2014J. Mullenbach et al.”Exploring affective communication through variable-friction surface haptics,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 3963-3972.CHI 2014I. Politis, S.A. Brewster and F. Pollick,”Evaluating multimodal driver displays under varying situational urgency,”in SIGCHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2014, pp. 4067-4076.CHI 2018A. Van Oosterhout, M. Bruns Alonso and S. Jumisko-Pyykko,”Ripple thermostat: Affecting the emotional experience through interactive force feedback and shape change,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 655.CHI 2018A.F. Siu et al.”Shapeshift: 2D spatial manipulation and self-actuation of tabletop shape displays for tangible and haptic interaction,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 291.CHI 2018P. Lopes et al.”Adding force feedback to mixed reality experiences and games using electrical muscle stimulation,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 446.CHI 2018H. Gil et al.”Whiskers: Exploring the Use of Ultrasonic Haptic Cues on the Face,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 658.CHI 2018J Gong et al.”Jetto: Using lateral force feedback for smartwatch interactions,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 426.CHI 2018P. Strohmeier, S. Boring and K. Hornbæk,”From pulse trains to coloring with vibrations: Motion mappings for mid-air haptic textures,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 65.CHI 2018S. Zhao et al.”Coding tactile symbols for phonemic communication,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 392.CHI 2018J. Bornschein, D. Bornschein and G. Weber,”Comparing computer-based drawing methods for blind people with real-time tactile feedback,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 115.CHI 2018Y. Zhao et al.”Enabling People with Visual Impairments to Navigate Virtual Reality with a Haptic and Auditory Cane Simulation,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 116.CHI 2018S. Je et al.”PokeRing: Notifications by Poking Around the Finger,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 542.CHI 2018H. Kim, C. Coutrix and A. Roudaut,”KnobSlider: Design of a Shape-Changing UI for Parameter Control,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 339.CHI 2018F. Kiss et al.”Navigation systems for motorcyclists: exploring wearable tactile feedback for route guidance in the real world,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 617.CHI 2018T. Komatsu et al.”Vibrational Artificial Subtle Expressions: Conveying System’s Confidence Level to Users by Means of Smartphone Vibration,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 478.CHI 2018J. Seo. et al.”Substituting Motion Effects with Vibrotactile Effects for 4D Experiences,”in CHI Proc. Conf. on Human Factors in Computing Systems. ACM, 2018, pp. 428.UIST 2018O. Schneider et al.”DualPanto: A Haptic Device that Enables Blind Users to Continuously Interact with Virtual Worlds,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 877-887.UIST 2018H. Pohl and K. Hornbæk,”ElectricItch: Skin Irritation as a Feedback Modality,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 765-778.UIST 2018M. Teyssier et al.”MobiLimb: Augmenting Mobile Devices with a Robotic Limb,”in UIST Proc. 31st Annual Symp. on User Interface Software and Technology. ACM, 2018, pp. 53-63.Table 8.8: The final included 24 HCI community papers used in our analysis.This represents24 out of 38 papers used for the final analysis.93Venue/Year Analyzed PublicationToH 2014M. Hirokawa et al.”Effect of haptic assistance on learning vehicle reverse parking skills,”IEEE Trans. on Haptics, vol.7, no.3, pp. 334-344, 2014.ToH 2014X. Chen et al.”Design of a robotic mobility system with a modular haptic feedback approach to promote socialization in children,”IEEE Trans. on Haptics, vol.7, no.2, pp. 131-139, 2014.ToH 2014J.C. Metzger et al.”Neurocognitive robot-assisted therapy of hand function,”IEEE Trans. on Haptics, vol.7, no.2, pp. 140-149, 2014.ToH 2014A. Theriault, M. Nagurka and M.J. Johnson,”Design and development of an affordable haptic robot with force-feedback and compliant actuation to improve therapy for patients with severe hemiparesis,”IEEE Trans. on Haptics, vol.7, no.2, pp. 161-174, 2014.ToH 2014A. Ajoudani et al.”Exploring teleimpedance and tactile feedback for intuitive control of the pisa/iit softhand,”IEEE Trans. on Haptics, vol.7, no.2, pp. 203-215, 2014.ToH 2014C. Pacchierotti et al.”Teleoperation of steerable flexible needles by combining kinesthetic and vibratory feedback,”IEEE Trans. on Haptics, vol.7, no.4, pp. 551-556, 2014.ToH 2018E. Olivieri et al.”Haptic feedback for control and active constraints in contactless laser surgery: concept, implementation, and evaluation,”IEEE Trans. on Haptics, vol.11, no.2, pp. 241-254, 2018.ToH 2018G. Korres et al.”A vibrotactile alarm system for pleasant awakening,”IEEE Trans. on Haptics, vol.11, no.3, pp. 357-366, 2018.ToH 2018G. Liu et al.”Effect of electrostatic tactile feedback on accuracy and efficiency of pan gestures on touch screens,”IEEE Trans. on Haptics, vol.11, no.1, pp. 51-60, 2018.ToH 2018T. Fukuda et al.”A pneumatic tactile ring for instantaneous sensory feedback in laparoscopic tumor localization,”IEEE Trans. on Haptics, vol.11, no.4, pp. 485-497, 2018.ToH 2018A.K. Han et al.”MR-compatible haptic display of membrane puncture in robot-assisted needle procedures,”IEEE Trans. on Haptics, vol.11, no.3, pp. 443-454, 2018.ToH 2018Y. Gaffary et al.”Toward Haptic Communication: Tactile Alphabets Based on Fingertip Skin Stretch,”IEEE Trans. on Haptics, vol.11, no.4, pp. 636-645, 2018.ToH 2018A. Teranishi et al.”Combining Full and Partial Haptic Guidance Improves Handwriting Skills Development,”IEEE Trans. on Haptics, vol.11, no.4, pp. 509-517, 2018.HS 2018W.H. Jantscher et al.”Toward improved surgical training: Delivering smoothness feedback using haptic cues,”in HAPTICS Haptics Symp. IEEE, 2018, pp. 241-246.Table 8.9: The final included 14 Haptics community papers used in our analysis. Thisrepresents 14 out of 38 papers used for the final analysis.8.4 Expert Ratings of Included and Excluded PapersAs mentioned in the Chapter 5 Focal Point 1 Results, we also visualized the expert ratings ofthe distributions of included and excluded papers. This was to help show what details pusheda community’s set of papers into being considered for haptic application design analysis.Each of the figures below will show these visualizations based on the different researchquestions associated with Focal Point 1. Different perspectives of the expert ratings will beshown (By Community, By Year, By Venue Size). Please note that there are some differencesin the visual language of the figures. Notably, compared with the figures in Chapter 5 FocalPoint 1 results, the following figures utilize side-by-side barcharts that showcases how bothcommunities fared in terms of details given for a specific area of haptic application design.Additionally, the sample information is not as detailed as the Chapter 5 figures, but they94are based on the same information presented in the Chapter 5 figures.Figure 8.2 represents how the communities behaved differently/similarly in haptic ap-plication design. The figure grouped all HCI venues against all Haptics venues used in ourstudy.95Figure 8.2: Haptic expert ratings of each community’s venues on different areas of hapticapplication design.96Figure 8.3 represents a baseline view (2014) in which future venues can be used to detectany changes in haptic application design patterns. Figure 8.4 represents the current view(2018) of community approaches to haptic application design. This view can be comparedagainst the 2014 view in Figure 8.3 to denote any changes in design patterns.97Figure 8.3: Haptic expert ratings of the 2014 community venues on different areas of hapticapplication design.98Figure 8.4: Haptic expert ratings of the 2018 community venues on different areas of hapticapplication design.99Figure 8.5 represents the community design patterns of the bigger sized venues (based onsample size). This view can be compared against the smaller sized venues in order to denoteany differences/similarities in design patterns. Figure 8.6 represents the community designpatterns of the smaller sized venues (based on sample size). This view can be comparedagainst the bigger sized venues to denote any differences/similarities in design patterns.100Figure 8.5: Haptic expert ratings of the bigger sample community venues on different areasof haptic application design.101Figure 8.6: Haptic expert ratings of the smaller sample community venues on different areasof haptic application design.102

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0389974/manifest

Comment

Related Items