ABSTRACT

Title of Dissertation: DESIGN CONSIDERATIONS FOR
REMOTE EXPERT GUIDANCE USING
EXTENDED REALITY IN
SKILLED HOBBY SETTINGS

Hanuma Teja Maddali, Doctor of Philosophy, 2023

Dissertation Directed by: Dr. Amanda Lazar
College of Information Studies,
University of Maryland

As compact and lightweight extended reality (XR) devices become increasingly

available, research is being reinvigorated in a number of areas. One such area for XR

applications involves remote collaboration, where a remote expert can assist, train,

or share skills or ideas with a local user to solve a real-world task. For example,

researchers have looked into real-time expert assistance [1] and professional training

of novices [2] in skilled physical activities such as field servicing and surgical training.

Even as our understanding of XR for remote collaboration in professional settings

advances, an area that has not been examined is how XR can support such expert-

novice collaboration in skilled hobby activities (e.g., gardening, woodworking, and

knitting). Metrics such as task accuracy or efficiency are often less important than

in professional settings. Instead, other dimensions, such as social connectedness and

emotional experience, may become central dimensions that inform system design.

In my dissertation, I examine how the XR environment can be designed to sup-



port the sharing of skills in hobby activities. I have selected gardening as a hobby

activity to examine remote skill-sharing in XR between experts and novices. Like

in other hobby activities, learning gardening practices remotely can involve asyn-

chronous, text, or image/video-based communication on Facebook groups. While

these may be helpful for individual questions, they do not capture the social, af-

fective, and embodied dimensions of gaining expertise as a novice through situated

learning in the garden. These dimensions can also be central to the experience of

the activity [3–7]. In my work, I seek to understand how to design a social XR

environment that captures these dimensions in ways that are acceptable and useful

to intergenerational expert-novice gardener groups.

Through my dissertation work, I answer the following research questions:

• How do practitioners of a particular hobby exhibit sociality and what kinds of

social interactions facilitate skill-sharing? What are some key opportunities

for computer-supported collaborative work in this space? [8]

• What are practitioners’ perceptions of using XR for skill-sharing? What are

the important dimensions of the design space and design scenarios for social

XR systems? [9]

• How do practitioners use different components of the activity space (e.g., tools

or sensory stimuli) and their affordances to facilitate social connection? What

context is essential to capture when reconstructing these objects virtually for

remote interaction in XR (e.g., interactivity and realism)? [9, 10]

• What are some design considerations for XR to support accessible interactions



that reflect the values and goals of an intergenerational group? [10,11]



Design Considerations for Remote Expert Guidance Using Extended
Reality in Skilled Hobby Settings

by

Hanuma Teja Maddali

Dissertation submitted to the Faculty of the Dept. of Computer Science of the
University of Maryland, College Park in fulfillment

of the requirements for the degree of
Doctor of Philosophy

2023

Dissertation Committee:
Dr. Amanda Lazar, Chair/Advisor
Dr. Paul T. Leisnham, Dean’s Representative
Dr. Matthias Zwicker
Dr. Huaishu Peng
Dr. Wayne G. Lutters



Acknowledgments

This dissertation work would not have been possible without the patience,

effort, and empathy of my participants, advisor, family, friends, and colleagues. It

has been a longer journey than I intended. But with their help, in some shape or

form, I feel I’ve become a more self-aware person and researcher compared to when

I started. Thank you.

ii



Table of Contents

Acknowledgements ii

Table of Contents iii

List of Tables vii

List of Figures viii

List of Abbreviations x

Chapter 1: Introduction 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.1.1 Skilled Activities and HCI: Why Gardening? . . . . . . . . . . 3
1.1.2 Learning How To Garden Remotely: Opportunities and Chal-

lenges for HCI and CSCW . . . . . . . . . . . . . . . . . . . . 4
1.1.3 The Potential and Challenges of XR in Skilled Hobbies . . . . 6

1.2 Research Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2.1 Study-1 (Published at ACM CHI 2020) . . . . . . . . . . . . . 8
1.2.2 Study-2 (Published at ACM CSCW 2022) . . . . . . . . . . . 9
1.2.3 Study-3 (Published at ACM CHI 2023 and Awarded Best Paper) 10

1.3 Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Chapter 2: Sociality and Skill Sharing in the Garden 13
2.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.3 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

2.3.1 HCI in the Outdoors . . . . . . . . . . . . . . . . . . . . . . . 17
2.3.2 Community Engagement in the Garden . . . . . . . . . . . . . 18
2.3.3 Gardeners’ Perspectives on Technology . . . . . . . . . . . . . 19

2.4 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.4.1 Study Procedures . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.4.2 Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.4.3 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.4.4 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

2.5 Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.5.1 Configuring Sociality . . . . . . . . . . . . . . . . . . . . . . . 25
2.5.2 Skill Sharing in the Garden . . . . . . . . . . . . . . . . . . . 33

iii



2.6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.6.1 Teaching and Learning Embodied Skills in the Garden . . . . 40
2.6.2 Designing for Varying Social Preferences . . . . . . . . . . . . 42
2.6.3 Negotiating Inclusion and Ownership . . . . . . . . . . . . . . 44

2.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.8 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

Chapter 3: Probing the Potential of Extended Reality to Connect Experts
and Novices in the Garden 49

3.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.3 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

3.3.1 XR for Remote Collaboration and Instruction . . . . . . . . . 55
3.3.2 Perspectives on Skill Acquisition . . . . . . . . . . . . . . . . . 57
3.3.3 Teaching and Learning in the Garden . . . . . . . . . . . . . . 59

3.4 Overview of Approach . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.5 Part 1: Exploratory Storyboards and Experience Prototypes . . . . . 63

3.5.1 Prototypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.5.2 Key Interactions Identified in Part 1: Instructing, Observing,

and Discussing . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.6 Part 2: Development of XR Prototypes . . . . . . . . . . . . . . . . . 71

3.6.1 Design and Technology Choices . . . . . . . . . . . . . . . . . 72
3.6.2 Activity 1 - Expert Tour . . . . . . . . . . . . . . . . . . . . . 74
3.6.3 Activity 2 - Awareness Building . . . . . . . . . . . . . . . . . 76
3.6.4 Activity 3 - Collaborative Garden Planning . . . . . . . . . . 77
3.6.5 Evaluation of XR Prototypes . . . . . . . . . . . . . . . . . . 79

3.7 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.7.1 Perspectives on Remote Gardening . . . . . . . . . . . . . . . 81
3.7.2 Skill Sharing Interactions Using the XR Prototypes . . . . . . 85
3.7.3 Connecting Interactions and Emotional Dimensions . . . . . . 90

3.8 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
3.8.1 Merits and Limitations of XR for Remote Skill Sharing in the

Garden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
3.8.2 Building Common Ground Through Connecting Interactions

in XR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
3.9 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
3.10 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
3.11 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Chapter 4: Understanding Context to Capture when Reconstructing Mean-
ingful Spaces for Remote Instruction and Connecting in XR 103

4.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
4.3 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

iv



4.3.1 Virtual Representation of Physical Spaces and Objects for Re-
mote Guidance in XR (RQ1) . . . . . . . . . . . . . . . . . . . 108

4.3.2 Intergenerational XR for Meaningful Informal Settings (RQ2) 112
4.4 Study Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

4.4.1 Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
4.4.2 Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
4.4.3 XR Remote Instruction Prototype . . . . . . . . . . . . . . . . 121
4.4.4 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

4.5 Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
4.5.1 Context Important to Users’ Experiences With the 3D Recon-

structed Activity Space in XR (RQ1) . . . . . . . . . . . . . . 127
4.5.2 Interplay Between Perception of XR and Prototype Usage in

Our Intergenerational Groups (RQ2) . . . . . . . . . . . . . . 135
4.5.3 Effect of Model Quality on Prototype Feedback as a Limitation137

4.6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
4.6.1 Capturing Context for Meaningful Reconstructions of Objects

and Spaces for Informal XR Learning . . . . . . . . . . . . . . 139
4.6.2 Reflecting on Privacy Nuances When Creating or Sharing 3D

Models for XR (RQ1C) . . . . . . . . . . . . . . . . . . . . . . 144
4.6.3 Capturing Context From the Garden for Social Connection

and Intergenerational XR . . . . . . . . . . . . . . . . . . . . 145
4.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
4.8 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

Chapter 5: Conclusion and Future Work 152
5.1 Key Takeaways and Contributions from Study 1 . . . . . . . . . . . . 152

5.1.1 Teaching and Learning Embodied Skills in the Garden . . . . 153
5.1.2 Designing for Varying Social Preferences . . . . . . . . . . . . 153
5.1.3 Negotiating Inclusion and Ownership . . . . . . . . . . . . . . 153
5.1.4 Accessibility and Connection . . . . . . . . . . . . . . . . . . . 154

5.2 Key Takeaways and Contributions from Study 2 . . . . . . . . . . . . 154
5.2.1 Merits and Limitations of XR for Skill Sharing in Gardening . 155
5.2.2 Augmenting Building Common Ground Through Connecting

Interactions in XR for Hobby Settings . . . . . . . . . . . . . 156
5.3 Key Takeaways and Contributions From Study 3 . . . . . . . . . . . . 157

5.3.1 Ideas on Supporting Meaningful Reflection During Remote
Instruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

5.3.2 More Implicit Context Required To Understand 3D Models
With Real-World Experience . . . . . . . . . . . . . . . . . . . 158

5.3.3 Perspectives on Privacy When Creating or Sharing 3D Recon-
structions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

5.3.4 Value of Reconstruction in Preserving Shared Memories . . . . 160
5.4 Reflections on the Research I Have Pursued as Part of This Disser-

tation Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
5.5 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

v



Appendix A:Storyboards used in Phase 1 166

Bibliography 168

vi



List of Tables

2.1 Self-Reported Participant Information (‘-’ indicates participant wished
to keep information private). . . . . . . . . . . . . . . . . . . . . . . . 22

3.1 Self-Reported Participant Information. . . . . . . . . . . . . . . . . . 61
3.2 Session Information. 1Due to time constraints, these participants

were unable to evaluate the awareness prototype. 4For sessions where
we were unable to recruit a novice, the first author (a novice) part-
nered with the expert. . . . . . . . . . . . . . . . . . . . . . . . . . . 62

3.3 Summary of interaction elements for the remote user of XR proto-
types in Study 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

4.1 Self-Reported Participant Information . . . . . . . . . . . . . . . . . . 115
4.2 Session Information. 1Participants alternated between mobile phones

(for onsite AR) and Oculus Quest headsets (when simulating remote
VR). 2P5 on Tablet (AR) and P6 on VR headset. 3P7 and P8 on
mobile, P9 on VR headset. . . . . . . . . . . . . . . . . . . . . . . . . 116

4.3 Examples of themes, subthemes, associated research questions (re-
lated theme), and underlying codes from data analysis. . . . . . . . . 126

4.4 Salient Findings and Implications . . . . . . . . . . . . . . . . . . . . 150

vii



List of Figures

2.1 P5’s drawing of his “secret garden” shows the fence (highlighted in
red) and a meditation hut (highlighted in blue). . . . . . . . . . . . . 24

2.2 P5 walking next to his fence covered with blackberry and honeysuckle
(left) and standing in his meditation hut (right). These created a
sense of privacy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.3 Drawing by P9. ADA-accessible teaching spaces are highlighted in
green and open to all visitors. . . . . . . . . . . . . . . . . . . . . . . 28

2.4 P1 decorated the garden for a national holiday with flags, visible
through the chain link fence, to show community sentiment. . . . . . 30

2.5 P9 showing the first author the difference between leaf venation of a
monocot (left) and dicot (right). . . . . . . . . . . . . . . . . . . . . . 34

2.6 Mock-up of a mobile screenshot showing how P3 asked for her friend’s
help with identifying a plant. . . . . . . . . . . . . . . . . . . . . . . 36

2.7 P8 used plastic sheeting after observing its usage by farmers. . . . . . 37
2.8 P1 pointing at deer skull (left). P3 holding deer antlers (right). . . . 38
2.9 Researcher’s 3D printed antler. . . . . . . . . . . . . . . . . . . . . . 39

3.1 Expert (on the right) using an iPad to observe video streaming from
a GoPro mounted on the head of the novice. The image shown here
is captured from that stream and shows the novice learning to use a
tool. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

3.2 Expert instructs a novice transplanting a plant, separated by a virtual
window (in red). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

3.3 Expert Tour design probe screenshot. The novice can draw or point
at objects or orient themselves with a map. . . . . . . . . . . . . . . . 75

3.4 Awareness design probe screenshot showing how the novice can cap-
ture photos using the camera tool. The photos preserve a view of
the environment for the novice to take note of and become aware of
changes over time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

3.5 Garden Planning design probe screenshot. In dialogue with the novice,
the expert selects a plant to place in the plot or draws garden bound-
aries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

3.6 Experienced gardener E14 (viewing the laptop) and novice gardener
N3 (wearing the Oculus HMD) dyad from session S8. . . . . . . . . . 79

4.1 Summarized timeline for each participant session. . . . . . . . . . . . 117

viii



4.2 Examples of meaningful objects and areas discussed during the garden
walkthrough. Participants focused on objects that might be meaning-
ful for their utility in activities they like, instructional value, aesthetic
reasons, etc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

4.3 One 3D model from the set of 3-4 models generated for each of the
eight study sessions. Models were generated from video clips provided
by participants of garden areas and objects they selected to be useful
for instruction or meaningful in some sense. The scale of these ranged
between large areas (e.g., butterfly garden) and small garden beds or
single objects (e.g., herb greenhouse). . . . . . . . . . . . . . . . . . . 119

4.4 Examples from session S7 of areas selected by participants along with
screen captures from devices used during Part 2 of the study. We can
see 3D models of the garden area generated and used for Scenarios 1
(rain) and 2 (sun) along with the simple sphere with dotted eyes for
user avatars. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

4.5 One of the models for session S2 which participants described as dif-
ficult to recognize and unnatural like a “Martian landscape”. Corre-
sponding photograph of the physical area in Figure 4.2b. . . . . . . . 137

A.1 Panels from the Collaborative Gardening storyboard. . . . . . . . . . 166
A.2 Panels from the Expert Mentor storyboard. . . . . . . . . . . . . . . 167

ix



List of Abbreviations

HCI Human Computer Interaction
CSCW Computer Supported Cooperative Work
DIY Do-It-Yourself
XR Extended Reality
AR Augmented Reality
VR Virtual Reality
MR Mixed Reality
HMD Head-Mounted Display
ADA Americans with Disabilities Act of 1990
CAD Computer Aided Design
WoZ Wizard-of-Oz prototyping

x



Chapter 1: Introduction

This thesis is centered around the design of remote collaboration systems for

practitioners of physically skilled hobby activities.1 Specifically, the goal of my

research is to understand how to design, build, and evaluate systems that enhance

the social and instructional experience of remote skill sharing in expert-novice groups

for hobby activities. The body of my dissertation is formed by three interconnected

studies with gardening as the hobby setting in which I have chosen to study remote

collaboration. Gardening is a useful case to explore design for collaboration in

hobbies as it has widespread intergenerational appeal and is viewed as a meaningful

activity for individuals, families, and communities [14].

Through an initial participant observation study (Study 1) with nine expe-

rienced gardeners, I identified opportunities and design considerations for socio-

technical systems in the activity space of the garden. The findings of this study

highlight the influence of practitioners’ social preferences as design considerations

when facilitating skill sharing interactions. Given the existing need for access to ex-

pertise in instructional programs, the findings also provided motivation for further

1In physically skilled activities, expertise is embedded in physical movements and a history of
interaction with materials [12]. Skilled hobbies can be defined as activities that privilege the joys
of production over the value of the product [13]. Examples of skilled physical hobbies include
gardening, woodworking, and soldering.

1



examination of remote guidance approaches.

I conducted this further examination of remote guidance by engaging 29 prac-

titioners in an iterative design and evaluation of three extended reality (XR) appli-

cations in Study 2. XR is a technology that is increasingly being studied and used

in remote expert guidance systems for skilled embodied activities in professional

(e.g., industrial design and surgical training) scenarios. Study 2 provides an un-

derstanding of the perception among practitioners of the applicability of XR as an

emerging technology for collaboration, specifically skill sharing, in informal hobbyist

scenarios. The study findings also indicate directions for XR to support affective

“connecting” interactions that help to build common ground [15] for learning be-

tween practitioners in these scenarios. Reconstructing and experiencing physical

spaces for practitioners as 3D models with AR/VR devices is one such direction.

It can be a powerful tool to provide environmental context for remote learning and

augment connecting interactions with emotional context (e.g., value seen by different

generations).

While there is much research that emphasizes sharpening sensory details of

the models, these only partially correlate with user needs and possibilities for infor-

mal learning with AR/VR. Study 3 identifies the social context that users expect

to capture when reconstructing personally meaningful physical spaces and objects

to learn and connect with family and friends. Study 3 also acts as a capstone for

my dissertation. It involves 1) the design of a multi-user remote XR system that

combines learnings from Study 1 and Study 2 and 2) a scenario-based evaluation of

the system in physical gardens by 18 gardeners in 8 intergenerational groups. The

2



findings from this study shed light on perspectives around creating and sharing 3D

models of meaningful objects and how they fit into the space of 3D contextual cap-

ture for instruction and connection. As part of this, I also highlight 1) implications

for authoring and sharing 3D content that can represent shared memories, and 2)

settings outside of gardens where exploring meaningful reconstruction of spaces and

objects could augment intergenerational learning and social interactions.

1.1 Background

1.1.1 Skilled Activities and HCI: Why Gardening?

The context of computer-supported collaborative work has expanded beyond

the workplace and into everyday lives and culture in the third wave of HCI [16].

Researchers are studying activities that occur in a diverse range of contexts, for

example, in the outdoors, in nature spaces, recreational spaces, and in learning

spaces for hobbies, crafts, and traditions. For activities that are embodied in nature

where using and sensing through one’s own body is important to learning and per-

forming the tasks, research has often focused on the ways that experts hone their

craft [17, 18].

In addition to developing individual expertise, however, practitioners are mo-

tivated to share their skills with friends, family, and others for broader goals that

impact their communities (e.g., encouraging self-reliance through DIY [19,20]). Cer-

tain activity sites (e.g., community gardens) foster interactions between people from

diverse backgrounds and serves as a site for the experiential learning of social and

3



civic skills. For my dissertation, I chose to focus on the activity of gardening to un-

derstand the role that technology might play in achieving the practitioners’ desired

social, personal, and community objectives related to skill sharing.

Gardening is an activity practiced by people across ages and demographics

whose impact can be felt at the personal as well as at the community level. However,

skills related to food growing have seen a significant decline due to the industrializa-

tion and mechanization of food production [21]. Researchers argue the importance

of reskilling in gardening and other craft practices as a way to return to authentic,

deep, and hands-on engagement with broader issues affecting the community [22,23].

Specifically, for food production, these broader issues include environmental sus-

tainability, food security, and social justice-oriented issues of increasing interest in

Human-Computer Interaction (HCI) and Computer-Supported Cooperative Work

(CSCW) [21]. Thus, researchers have specifically called for further study on how

technology might support education around food production [4]. My work studies

remote expert guidance approaches that attend to these broader motivations.

1.1.2 Learning How To Garden Remotely: Opportunities and Chal-

lenges for HCI and CSCW

Though there is certainly value in studying in-person skill sharing in garden-

ing, this dissertation focuses on the topic of remote skill sharing for several rea-

sons. First, past research has pointed out that the lack of local access to gardening

knowledge, volunteers, and community support challenges the implementation of

4



instructional gardening programs [24]. There is thus a need to understand whether

remote skill sharing can yield benefits for those who lack access to local experts.

Second, there are many open research questions around whether skill sharing in

embodied, sensory-dependent activities can translate online. In-person learning of

these activities is effective for a range of reasons, including the ability to provide

tailored feedback to a novice in complex and subjective parts of a task [25] and

to introduce the novice to the craft culture and vocabulary by contextualizing it

through actions (e.g., pointing to features of an artifact or picking up the right

tool) [12]. Specific to sharing gardening skills, past research indicates that novices

gain skills and an understanding of the community culture through immersion in

the garden environment with experienced practitioners [3, 26]. However, existing

remote approaches largely take an asynchronous, non-immersive approach to share

skills on a large scale [5, 6]. For example, people can use online forums to share

pictures of their plants and ask questions (e.g., on Reddit)2, or watch step-by-step

guides where experienced gardeners share techniques (e.g., on YouTube). These ap-

proaches lack some of the dimensions that are central to gardening interactions: an

immersive sensory experience [4,6] and social intimacy when interacting on-site with

other practitioners [4,5,27]. My work examines whether XR systems might support

these social and sensory dimensions and if they might be accepted by practitioners

for this purpose.

2https://www.reddit.com/r/vegetablegardening/

5



1.1.3 The Potential and Challenges of XR in Skilled Hobbies

Collaboration using XR systems is an active area of research and development.

There is much interest in the HCI community in understanding design considera-

tions for remote guidance in embodied skilled activities. More recently, as compact,

lightweight XR devices become increasingly available, research is being reinvigorated

in several areas. One such application involves remote guidance, where a remote

expert can assist, train, or share skills or ideas with a local user to solve a real-world

task. For example, researchers have looked into real-time expert assistance [10] and

professional training of novices [9] in skilled physical activities such as field servicing

and surgical training. The capability of professional and consumer XR devices to re-

construct objects and spaces as 3D models for remote collaboration is also becoming

increasingly powerful and ubiquitous. Even as our understanding of XR for remote

collaboration in professional settings advances, an area that has been unexamined is

how XR can support such expert-novice collaboration in skilled hobby activities. In

skilled hobby activities, such as gardening, woodworking, or knitting, professional

goals such as task accuracy or efficiency may not be the parameters around which

to build the system. Through my work, I uncover other dimensions, such as social

connectedness and affect (i.e. the emotional experience), that are central dimensions

to inform system design for skilled hobby activities.

6



1.2 Research Approach

In my dissertation, I examined how remote collaboration through XR can

be designed to support the sharing of skills in physical hobby activities such as

gardening. I sought to understand the perceived role of technologies such as XR

and design approaches that capture useful dimensions for hobby practitioners to

share their skills in ways that are acceptable to expert-novice gardener groups in

informal and intergenerational scenarios.

My dissertation research involved three studies in sequence. In my first study,

I explored individual and social experiences in the garden activity space and identi-

fied design considerations for skill sharing with socio-technical systems. The findings

from Study 1 have been published at ACM SIGCHI Conference on Human Factors in

Computing Systems (CHI 2020) [8]. Using these design considerations from Study 1,

I iteratively went from low to high-fidelity designs for remote XR prototypes to sup-

port skill sharing with feedback from groups of gardeners in Study 2. The evaluation

of these prototypes by expert-novice dyads allowed me to understand perceptions of

remote learning using XR and the potential for XR to augment social “connecting”

interactions in informal learning settings. The findings from Study 2 have been pub-

lished at ACM SIGCHI Conference on Computer-Supported Cooperative Work and

Social Computing (CSCW 2022) [9]. In Study 3, I applied learnings from Studies

1 and 2 to build a remote collaboration XR system for an intergenerational setting

that could be evaluated in a naturalistic setting. I investigated learning and con-

necting experiences specifically with reconstructions of objects and spaces in gardens

7



that were meaningful to closely-related intergenerational groups of gardeners. The

findings from Study 3 have been published at ACM SIGCHI Conference on Human

Factors in Computing Systems (CHI 2023) [10].

1.2.1 Study-1 (Published at ACM CHI 2020)

This study laid the groundwork to understand sociality in the garden and how

it influences skill sharing interactions between practitioners. I conducted participant

observation sessions with nine experienced gardeners at their gardening sites and

qualitatively analyze the data from the sessions to address the following research

questions:

• How do practitioners engage socially as they practice their hobby?

• What kinds of social interactions facilitate skill sharing in the garden?

• What are some key opportunities for computer-supported collaborative work

in this space?

The findings of the study highlighted design considerations for teaching and learning

in the garden space based on how practitioners engaged in collocated or remote social

interactions with others in their community. I found that learning occurs through

collocated interactions inside the garden, using a range of digital resources outside

the garden, and even indirectly by observing other practitioners’ gardens. Based

on my findings, I identified opportunities for remote approaches to skill sharing,

including video telepresence and XR. In Study 2, I examined how one of these

8



types of technologies, XR, can facilitate the immersive and embodied interactions

that practitioners associated with gardening. A topic I return to in Study 3, based

on findings from Study 1, is ensuring that digital approaches are inclusive of the

intergenerational and mixed-ability groups who benefit from gardening.

1.2.2 Study-2 (Published at ACM CSCW 2022)

In this study, I determined the feasibility and reveal considerations for remote

approaches for learning in expert-novice groups in the garden. From Study 1, XR

appeared to be a promising approach to support embodied interactions and immer-

sion in the remote activity environment. I pursued this direction by conducting

user-centered design sessions with 29 gardeners in three phases: obtaining feedback

from participants on low-fidelity prototypes, developing XR design probes based on

feedback on the prototypes, and finally evaluating the XR probes. The research

questions addressed in this study include:

• What are practitioners’ perceptions of using XR for skill sharing?

• What are the important dimensions of the design space and design scenarios

for skill sharing using social XR systems?

Through this study, I identified ways that XR systems for skill sharing could aug-

ment interactions that connected practitioners with each other’s motivations and the

activity space. This included visualizing community (e.g., environmental) impact

and communicating how experts measured by using their senses. In my findings,

I also found a pervasive theme of connecting through affective connecting interac-

9



tions. These interactions included sharing stories about objects in the garden, often

with the broad goal of passing on a love for the activity. The objective of designing

to connect differentiates this work from existing research on remote expert guidance

that uses XR for professional settings (e.g., surgery) and from prioritizing result-

oriented outcomes (e.g., accuracy and completion time). These findings motivated

the research focus of Study 3 to further understand how connecting interactions in

hobby learning settings can be augmented using XR.

1.2.3 Study-3 (Published at ACM CHI 2023 and Awarded Best Pa-

per)

My third study derives from the previous two studies, where XR that was cen-

tered around connecting activities appeared to be a feasible approach to remote skill

sharing in hobby settings. Additionally, I returned to two needs I identified in the

previous studies. The first is a need from Study 1 to design for an intergenerational

setting by involving and designing for the values and constraints of older gardening

experts. Second, a finding from Study 2 was how interaction with virtual objects and

remote environments in XR affected connection (e.g., sharing flowers and stories) as

well as instruction (e.g., measuring by observation). The ability to work with and

share virtual replicas of physical objects (e.g., virtual tools, plants, or garden plots)

was found to be a key advantage of the XR approach over conventional video chat,

especially for connecting interactions. However, there is a lack of understanding of

the social affordances of the objects in the hobby practitioners’ activity space and

10



how that would translate to virtual replicas in XR environments. In Study 3, I built

and deployed a high-fidelity XR system in the naturalistic setting of participants’

own gardens. I evaluated the system through observation of participants’ system

usage, and semi-structured interviews to evaluate social presence and connection

with mixed-ability inter-generational groups.

The objective of this study was to expand our understanding of how to support

connecting and instructional interactions between inter-generational, mixed-ability

groups using virtual representations of objects that were meaningful to users in the

remote activity environments in XR. The research questions addressed through this

study include the following

• What is the important social and sensory context that can influence infor-

mal XR users’ remote experiences with virtual representations of real activity

spaces and objects?

– What context is important to users for learning experiences with recon-

structed spaces and objects?

– What context is important for connecting interactions with reconstructed

meaningful spaces and objects?

– What context is important to users when sharing reconstructed mean-

ingful spaces and objects?

• What are design considerations for XR to support the values and goals of an

informal inter-generational group interacting with the virtual representations?

11



1.3 Organization

This thesis is organized as a three-paper thesis. Chapters 2-4 cover a de-

scription of the methods, findings, and discussion points from Studies 1, 2, and 3

respectively. Chapter 2 consists of my article published at ACM CHI 2020 [8] from

my participant observation study (Study 1). Chapter 3 consists of my article pub-

lished at ACM CSCW 2022 [9] for the iterative low to high-fidelity prototype design

study (Study 2). Chapter 4 consists of my article published at ACM CHI 2023, for

the scenario-based evaluation of working with reconstructions using my final remote

XR system (Study 3). Each chapter will include subsections that describe in detail

the motivations, background literature, methods employed, findings, design impli-

cations, and limitations of the studies. Chapter 5 concludes this dissertation with

a summary of Studies 1-3 along with my contributions to the field of HCI and XR

design and reflections on my learnings through this thesis.

12



Chapter 2: Sociality and Skill Sharing in the Garden

2.1 Overview

Understanding the existing experiences of practitioners of skilled hobby activi-

ties, their relationship with their activity space and other practitioners, and dynam-

ics with technology is crucial for designing effective technology that enhances their

activities. In the first study of my dissertation, I investigated the role that collab-

orative and social computing technologies might play for practitioners of gardening

as one such hobby activity. Gardening is an activity that involves several dimen-

sions of increasing interest to HCI and CSCW researchers, including recreation,

sustainability, and engagement with nature. I conducted participant observations

with nine experienced gardeners aged between 22 and 71. Through this process,

I discovered that gardeners constantly adapt their surroundings to accommodate

their preferences for social interaction. They share their physical skills and assist

others in tuning into sensory information in person, but also use the features in

their garden to facilitate learning for others who observe them and their spaces.

From these findings, I discuss the concept of sociality within gardening and iden-

tify design considerations for skill sharing using collaborative technologies in this

context. These design considerations also guided my subsequent research focus on

13



expert-novice and informal learning interactions in studies 2 and 3.

This chapter on Study 1 is adapted from my paper on “Sociality and Skill

Sharing in the Garden” [28] published in the proceedings of the 2020 ACM SIGCHI

Conference on Human Factors in Computing Systems. As the first author, I led

the study design, data collection, analysis, and written and oral presentation of this

work.1

2.2 Introduction

HCI researchers are examining outdoor activities and nature spaces as sites

of recreation, learning, and social interaction. Gardening is one such outdoor ac-

tivity with ties to other topics of interest in HCI such as food sustainability and

civic engagement with environmental issues. Gardening is an activity that builds

community and increases residents’ attachment to their neighborhood [29]. It fos-

ters interactions between people from diverse backgrounds and serves as a site for

the experiential learning of social and civic skills such as leadership, community

organizing, and cultural competency [29,30].

People are also drawn to gardening for opportunities to spend time outdoors

and with family [14]. A 2014 report estimates that one in three households in

the United States engage in food gardening, or urban agriculture [14]. These 42

million households include people of all ages. Yet, not all groups have equal access

to getting engaged in gardening. Access to land for gardening in an urban setting

1My co-author, Dr. Amanda Lazar, provided feedback and suggestions throughout the project
and paper write-up.

14



is limited [31], a lack of gardening knowledge can limit participation [24], and for

some groups with mobility constraints, technology typically focuses on indoor living

rather than outdoor spaces [32,33]. With a growing interest in the opportunities it

presents for technological applications for group interaction, and in the work still

needed to promote access to this activity, gardening appears to be an area ripe for

HCI research.

However, when considering design in the garden space, it is required to turn

to the body of past research that has found that practitioners are sensitive to how

technologies are introduced in the garden space. For example, automation using

sensor networks can be perceived as obstructing the sensory, embodied, emotional

feeling of engaging directly with nature [4,6,27]. Gardeners may trust their own lo-

calized, developed knowledge over scientific models [34]. Poorly designed technology

can also impede important social practices in the garden, for example, the trans-

mission of skills from experienced to novice gardeners [6]. The drawbacks of purely

technological approaches when engaging gardeners, paired with the social nature of

gardening, point towards exploring social-computing design approaches as a fruitful

area of research [5, 35]. Understanding the potential role of social technologies in

this space requires an examination of where sociality exists in the garden, as well as

the particular kinds of interactions that might need to be supported.

Our research takes, as a starting point, findings from past work: that garden-

ing is sensory and emotional, with social practices that have been built around these

activities over time. Given that past work has typically sought gardeners’ perspec-

tives on technologies that transmit information to gardeners (e.g., soil quality [36,37]

15



and temperature [35] sensors), we return to the garden setting with an ethnograph-

ically informed approach, engaging in participant observations with nine gardeners

to identify opportunities for HCI. Our paper offers three contributions. First, we

provide an understanding of sociality in terms of where it exists in the garden and

how gardeners configure desired levels of social interaction, for example, through

physical arrangements such as letting vines grow over a fence to obscure the view of

those passing by. Second, we highlight skill sharing as a key domain for social design

in this space. We find that this process of skill sharing is supported through different

levels of engagement with practitioners: in addition to direct interaction and the use

of digital platforms, practitioners learn techniques and other information through

the observation of others’ gardens. Finally, we contribute design considerations for

collaborative technologies in outdoor settings with implications for embodied skill

sharing and inclusion. We suggest that technology designed for social computing in

the garden should balance respecting dynamic sociality preferences with motivating

community engagement and collaboration.

2.3 Related Work

Below, we discuss research on technology for outdoor activities in nature

spaces, the garden as a site for community engagement, and perspectives on tech-

nology in the garden.

16



2.3.1 HCI in the Outdoors

Recent HCI research situates technology design in a range of outdoor activi-

ties. These include recreational activities (e.g., tourism [38], scuba-diving [39], and

paragliding [40]), fitness (e.g., running [41], wall-climbing [42], and cycling [43]),

and nature activities (e.g., hiking [44], foraging [45], and monitoring wildlife [46]).

Design in this space has considered motivations to engage outdoors as well as the

value that technology brings to personal and shared experiences. Some research on

outdoor activities draws on social facilitation theory [47] to support social interac-

tion in fitness groups through revealing information such as the speed of runners [41]

and heart-rate of cyclists [43] to others. These studies discuss insights from in-situ

presentations of individual and group performance metrics and their potential for

supporting group togetherness and motivations for fitness.

In contrast to work that supports social experiences around outdoor activities,

other research helps people disconnect from other people and become more immersed

in nature. For example, research has supported purposeful solitude in nature by

informing individuals of nearby hikers [44]. With the rising interest in HCI in

the outdoors, it is important to understand what design considerations exist for

supporting sociality in outdoor spaces. In our work, we find the importance of

acknowledging varying, rather than static social preferences in outdoor spaces —

a concept that has previously only been considered in a traditional indoor office

setting. We discuss how designing for learning or skill sharing in the garden space

should account for these varying social preferences.

17



Nature spaces are another domain of interest in outdoor HCI, including re-

search that explores the role of technology to support collaboration. Research has

examined the design of collaborative technology for search and rescue teams try-

ing to maintain situational awareness in wilderness [48] and to support simulations

in high-risk outdoor recreation areas [49]. Nature spaces also provide an avenue

for environmental learning. For example, Soro et al. describe an IoT “Ambient

Birdhouse” designed to interest children in engaging with nature by becoming more

aware of bird calls and discuss how it could be used as a catalyst for learning and

socializing [50]. Liu et al. designed three wearables for mushroom foraging to

“offer a vision of wearables extending our human sensory capacities into the en-

vironment” [45]. This vision offers people the capacity to “notice, attend to, and

become struck by nonhuman lives” and, in the case of one of the prototypes, also

share the information that they gather with others [45].

2.3.2 Community Engagement in the Garden

A common approach in sustainability HCI research is to design for community

engagement to encourage strong civic activity around pro-environmental goals. For

example, YardMap supports professionals and citizen-scientists in mapping personal

carbon-neutral yard practices, learning about their local environment, and discussing

their potential impact on habitats [51]. The inclusion of people from different cul-

tural backgrounds [52] and expertise levels [51] through knowledge-sharing [5] and

capacity-building [4] is seen as an important mechanism for building sustainable

18



communities.

Jrene Rahm’s work, centered around an inner-city youth gardening program,

highlights the role of active social participation in creating opportunities for devel-

oping expertise as a novice [26]. Novices, through situated learning, gain skills and

an understanding of the community culture through their interactions with peers

or more experienced practitioners and immersion in the garden environment [3,26].

The ways that gardeners become experts in using their senses to notice and observe

lead to opportunities to support people in new ways of engaging with the world

and with other practitioners towards more sustainable futures [6, 27, 53]. However,

researchers have also noted the tensions that arise when designing for engagement

in communities with diverse expertise levels. For example, managing the territo-

rial behaviors of experts is important in encouraging participation from novices and

allowing them to develop a feeling of attachment and ownership towards the com-

munity [54]. In our paper, we acknowledge the value of garden spaces in cultivating

a sense of community and engaging with other practitioners for skill sharing. We

discuss this in light of the tensions that we find in encouraging community inclusion

while maintaining ownership of the garden space.

2.3.3 Gardeners’ Perspectives on Technology

Several studies have focused on gardeners in community and residential set-

tings. This past research has found that gardeners do use many digital tools, often

to share information or support coordination. For example, Wang et al. analyze

19



collaboration between gardeners, finding that they use different tools for informa-

tion and knowledge sharing as well as scheduling work activities [55]. In a study of

handwork, Goodman and Rosner note that though gardeners and knitters use many

different digital tools, they define their values in opposition to stated negative char-

acteristics of technology [27]. For example, being engaged rather than disconnected

means “committing to the material details of making objects oneself,” rather than

using a system to cut oneself free of the task of watering [27].

A common theme of the body of work on gardening is the tensions that arise

with technology. For example, Lyle et al. highlight how gardeners learn through

experimentation and observation, as well as the importance of sharing knowledge

between community members [5]. In this context, Baumer and Silberman present

sensor nets for data-driven gardening as a case study of when the implication is

not to design a particular technological solution [35]. Sensing systems to support

automation of tasks or decisions (e.g., automatic watering based on soil moisture)

appear to be viewed negatively by gardeners in much past research, as they are

seen as interrupting existing values and processes, such as direct interaction between

gardeners and plants [4,6] that help gardeners develop environmental knowledge and

intuition [6]. Further, automation can interrupt the transfer of knowledge between

senior and novice members of the community [6].

This past literature lays the groundwork for investigating collaborative tech-

nologies and social computing in community and residential gardening as a fruitful

area of investigation, in that it matches the social and collaborative nature of gar-

dening and moves away from purely technological solutions [35]. Understanding the

20



potential role of collaborative technologies requires filling a gap in our understand-

ing of how sociality manifests in the garden. In our paper, we discuss our findings

on two such components of sociality: how a practitioner’s sociality preferences are

reflected in their working space, as well as the level of active engagement with other

practitioners when teaching and learning skilled activities in the garden.

2.4 Methods

Fieldwork was conducted over the summer (June through September 2018)

in the Mid-Atlantic region of the United States. Below, we describe the study

procedures, participant information, and our analysis.

2.4.1 Study Procedures

We took an ethnographically informed approach to data collection and anal-

ysis [56]. Sessions involved a 15-minute interview, a brief drawing prompt, and a

60-minute participant observation session. The interview included questions such

as participants’ motivation, frequency of gardening, and self-described level of ex-

pertise. In the drawing prompt, participants drew the physical sites where they

gardened, including the places that held meaning as well as where they grew dif-

ferent plants and kept tools. In the participant observation, gardeners were asked

to engage in activities that they normally would do around their garden. The first

author shadowed and worked alongside gardeners and asked questions when relevant

to the task at hand. This included, for example, questioning how gardeners made

21



ID Age Gender Ethnicity Observation Site

P1 60 Male - Public Community Garden
P2 26 Female White Home Garden
P3 61 Female White Home Garden
P4 37 Male Indian University Community Garden
P5 - - - Private Community Garden
P6 - - - University Community Garden
P7 37 Female White University Community Garden
P8 71 Male African Public Community Garden
P9 22 Female White Shade Garden (Ornamental)

Table 2.1: Self-Reported Participant Information (‘-’ indicates participant wished
to keep information private).

particular decisions.

Data collected included the sheet from the drawing prompt, observation notes,

video, and audio recordings. The video was collected with a head-mounted GoPro

Hero 5 action camera. Our sessions yielded approximately 800 minutes of audio and

video recordings (93 minutes per session on average). In parallel with data collection,

both authors spent time becoming familiar with the process of gardening. The

first author participated in weekly volunteer sessions at the university community

garden for four months, and the second author engaged in gardening in her backyard.

These experiences informed the study protocol and our understanding of gardening

practices and the process of learning gardening skills.

2.4.2 Participants

Nine participants who self-identified as gardening regularly were recruited

through local community garden e-mail lists, fliers posted on campus, word of

mouth, and snowball sampling. Participants were between the ages of 22 and 71 (

average=45 years, std. dev=19.3 years). All sessions involved participant observa-

tions, but four individuals did not engage in the initial 15-minute interview (P5, P6,

22



P7, and P9) and two did not engage in the drawing session (P1, P6) due to time

constraints.

We attempted to recruit from a range of gardening configurations to under-

stand how experiences may vary in different spaces. Participants gardened in dif-

ferent arrangements, from private backyards to public community gardens2 (Table

2.1). All participants grew food items, the most common being tomatoes, peppers,

and herbs. Almost half of the participants grew flowers for themselves, and the

majority grew flowers for pollinators. During the participant observation, we asked

individuals to engage in whatever tasks they might naturally be doing that day. This

ended up including a variety of activities: weeding, watering, trellising, harvesting,

decorating, and just relaxing in the garden.

2.4.3 Analysis

Our constructivist grounded theory approach to analysis [57] was as follows:

the first author open-coded two transcribed interviews and three sets of observation

notes to create a preliminary set of codes and emerging themes. The research

team met to discuss these codes over several sessions and became interested in

themes relating to Sociality (with codes such as “being accessible to passersby,”

“having informal boundaries in the shared plot,” and “viewing the garden from

an outsider’s point-of-view”) and Skill Sharing (with codes such as “learning from

someone who seems more experienced,” “observing decorations on neighbors plot,”

2A community garden is a single piece of land gardened collectively by a group of people. Table
2.1 indicates whether these gardens were situated on private or public land. Public community
gardens are usually managed by the local government.

23



and “sharing a photograph to describe plant condition”). The first author then

coded the rest of the transcribed interviews for these themes, adding additional

codes as they emerged. The research team related codes to each other through an

iterative process of memoing and theorizing, engaging in constant comparison of

data to understand and refine a set of high-level themes.

Figure 2.1: P5’s drawing of his “secret garden” shows the fence (highlighted in red)
and a meditation hut (highlighted in blue).

Figure 2.2: P5 walking next to his fence covered with blackberry and honeysuckle
(left) and standing in his meditation hut (right). These created a sense
of privacy.

24



2.4.4 Limitations

Though participants were diverse in the range of settings in which they gar-

dened, the small number of participants, all from the US, and our approach to

recruitment and analysis means these findings cannot be generalizable. The em-

phasis on skill sharing that arose was likely shaped by our method of participant

observation and participant roles in gardens (e.g., working in a community learn-

ing garden). Future work is needed to examine a more diverse and comprehensive

sample.

2.5 Findings

In this section, we discuss where sociality arises in the garden and the ways

gardening skills are taught and learned. We find that gardeners configure desired

levels of sociality. One way they do so is through physical arrangements. Elements

such as the type of fencing (e.g., honeysuckle-covered fences or a chain link fence)

can indicate ownership and manage interaction with other gardeners and passersby.

One kind of interaction that takes place in the garden is the learning and teaching

of skills between gardeners. Skills are shared directly, and also indirectly through

observation of others’ gardens.

2.5.1 Configuring Sociality

Participants engaged in different levels of sociality. This appeared in the actual

ways they went about gardening: P5 intentionally gardened alone, P2 worked with

25



her partner on most major tasks, and P4 liked to engage with and learn from others

in his community garden. P3 touched on the ways that different gardeners might

be drawn to different kinds of gardening arrangements. Where she gardened, it was

“close, you share a lot of space... people that want to be on their own, they wouldn’t

come here.” Many gardeners, though, were not solely social or private gardeners –

they chose to be private or social depending on their mood or the activity they were

doing.

2.5.1.1 Managing Interactions With Other Gardeners

Gardeners use physical features of gardens, such as raised beds and hedges, to

support desired types of social interactions. P1 explained that where he gardened,

plots involved “raised beds with wood around them so we each know our boundaries.”

P5 configured his space by letting vines grow on top of existing separations of plots

in his community garden to get more privacy: he pointed to the fence on his plot

(Figure 2.2) “where all the honeysuckle grows... [the] privacy gives me the secret

garden feeling that I like.” He appreciated being alone in the garden to find space

for introspection, meditation, and the feeling of getting away from culture. P5’s case

shows a gardener using natural and built features to create a more private space in

a community garden. In a contrasting example, P3 sometimes shared tasks with her

neighbors in a backyard that included both their garden spaces. It made a private

space more social by bringing together “people that are okay being close to other

people.”

26



In P5’s example, letting honeysuckle grow wild created a desirable social ar-

rangement for him. P9, on the other hand, spoke about how she arrives at a socially

desirable space by picking up debris – though she leaves leaf litter and smaller sticks

as a way of “keeping it natural”. P9 clears the pathways in the garden of debris to

make it “functional for everyone,” including those with disabilities. She explained

that part of the community learning garden where she worked was designed to com-

ply with the Americans with Disabilities Act3, “So people in wheelchairs or with

disabilities can easily access this part of the garden, and with the raised beds they

can participate in gardening just as well [as] people who don’t have a disability”

(Figure 2.3). P9 appreciated “how well this space includes a wide range of people.”

Like in the example above, physical configurations (in this case, the removal of

large debris) are ways that gardeners reach desired levels of sociality not only based

on personal preference but also based on policy or community-wide decisions. Some

community gardeners talked about how garden managers played a role in enforcing

community rules of particular sites. According to P6, managers resolved issues

with a plot that could affect other members, such as directing members to remove

weeds that could spread to other plots. Managers also disseminated news related to

group activities, such as putting down wood chips on paths. Even when gardeners

were coordinated to support adherence to policies about community space, privacy

preferences could be preserved. P1 explained that “the manager can send a message

to all of us ... we see there’s a list of emails, but we don’t know which necessarily

3The ADA is a civil rights law that prohibits discrimination against individuals with
disabilities in all areas of public life, including jobs, schools, transportation, and all pub-
lic and private places that are open to the general public. For more information, see
https://www.dol.gov/general/topic/disability/ada.

27



Figure 2.3: Drawing by P9. ADA-accessible teaching spaces are highlighted in green
and open to all visitors.

from the address refers to which person. [We] certainly don’t know which person

refers to which plot.”

2.5.1.2 Interaction With Those Outside the Garden

Participants also configured social interactions not only for fellow gardeners

but also for the broader community that comes into contact with gardens both

directly and indirectly. In Heitlinger et al.’s study, a central value of a farm garden

is inclusion [4]. In our study, we found that a commitment to inclusion shared

28



by many gardeners was in tension with preferences to create divisions between the

garden and the outside world. Participants saw boundaries that they created or that

were features of the space not only as important to keep out animals and people

who might take produce or flowers, but also to create a sense of privacy. P3 enjoyed

seeing passersby who would complement her flowers, saying that, “[it] is very nice

because you see people coming, passing by ... They are far away enough that they are

not in your space...”. Her garden was separated by an informal boundary created

by elevation from the passersby in a shared green space.

Though these separations were important for gardeners to achieve a level of

social interaction that was desirable for them, they did think about the ways that

some barriers to the outside world might come off as uninviting and worked to cre-

ate a more welcoming space without necessarily letting others into the garden itself.

P1’s plot in the community garden had a wire fence with a lock on it. When asked

if the garden saw visitors, perhaps children and their parents, from the bordering

playground he mentioned that the fence might have unfortunately created a feeling

of exclusion for the community: “There’s a sense, perhaps because of the fence,

that the gardeners want to be left alone and outsiders don’t bother [with them].” P1

discussed how he decorates his garden, for example, with flags for the US holiday

of July 4th (see Figure 2.4), to show “community sentiment.” This was so “people

outside the fence and [who] can’t get in might feel a little less excluded and maybe

it’s good public relations for the garden.” P1’s chain-link fence allowed individuals

to see the decorations he placed in his garden. P8 described an arrangement outside

a community garden that encourages community inclusion while preserving bound-

29



aries. Benches were arranged just outside the fence, in a way that invited outsiders

to sit and observe the garden.

Figure 2.4: P1 decorated the garden for a national holiday with flags, visible through
the chain link fence, to show community sentiment.

While some participants, like P7, described gardening as an opportunity to

“disconnect from a lot of technology,” smartphones and social media played a visi-

ble role when managing interactions with people outside the garden as well as those

in its physical proximity. Participants (P1, P3, P4, P6, P7, P8, and P9) frequently

mentioned capturing photographs using their smartphones and sharing them, for

example, on Facebook and Instagram. These were usually used to enable a pas-

30



sive form of interaction through the posted photographs and other content like, for

example, comments that compliment the garden (P1, P4, and P8).

Gardeners also described encouraging active involvement with the garden

through volunteering opportunities (P6, P7, and P9). P4 describes one such activity

where he posted online to invite others to make and share a sauce with peppers.

In general, we also find that participants described interactions of a mostly positive

nature (P1, P3, P4, and P8), both online and offline. P1 explains that it’s “a polite

thing” since “people get sensitive in the garden. Something about the place, some-

thing about the activity, you very much want to hear praise.” Though social sharing

was an important usage of photographs, photographs were also sometimes taken to

keep a personal record of the garden’s progress over time (P1, P3, and P8). For

example, P3 described taking, “photos of the flowers, because it’s fun to remember

when they bloom or just they are beautiful.”

2.5.1.3 Cultivating Desired Emotional States and Relationships

In addition to providing or preventing others from access to gardening spaces,

gardeners also used physical arrangements to cultivate certain emotional states for

themselves. P2 and P3 placed chairs in or within view of their gardens and spent

meals and time with their partners enjoying the ambiance. Two gardeners described

feeling meditative when gardening, with P5 reserving a space for meditating in the

garden – his “little meditation hut” (see Figure 2.1). These findings are consistent

with work from anthropology discussing boundaries within the garden itself, where,

31



“there are also separations between different areas and particular functions and

activities associated with each” [58].

Plants and objects in the garden also became ways that participants connected

with others outside the garden. Over the course of the study, the first author was

offered the following items from participants: beans, tomatoes, eggplants, peppers,

strawberries, ground cherries, basil, and three types of flowers – zinnia, globe ama-

ranth, and ageratum. A few gardeners grew items specifically to give as gifts, such

as P2 who grew catnip for her friend’s cat. Past work has also found that gardeners

share produce with friends or fellow gardeners [4, 27]: we find that gardeners also

gave away produce to benefit the community at large. P7 explains, “We do harvest

a lot of stuff and donate it to the campus pantry ... because 15 percent of our student

population is food insecure.” P4 and P8 shared produce with community members

at their place of worship. P4 describes they do this in part because “it helps save

the [place of worship] some money.”

In addition to connecting to other individuals or a broader community in the

present, participants used gardening to reinforce feelings of connection to people or

places from the past. P8, who grew hot peppers on one of his plots, referred to how

people from his native country love those peppers. P3 described how her hellebore

plant “reminds me of my mom, because she always had them,” and her gardening

toolbox housed a tool that reminded her partner of his father. She called these

“memory objects.” These examples highlight how the garden is shaped to create a

personal space that reflects the gardener’s relationship with a community or loved

one.

32



2.5.2 Skill Sharing in the Garden

A form of interaction that we discuss in this paper is skill sharing in the garden,

a recurring theme in our interviews and observations. More experienced gardeners in

our study often had formal and informal teaching roles, but even a master gardener

such as P7 acknowledged that “you never stop learning in this job, which is one

of the other reasons why it’s so enjoyable.” We detail the different forms in which

knowledge and skill sharing about the sensory-rich, embodied practices of gardening

took place.

2.5.2.1 Tacit Knowledge Communicated in Co-Located Learning

In past work, being physically collocated with other gardeners allowed novices

to get help from more experienced individuals, for example when dealing with slugs

[55]. Our findings reveal that co-located gardening enables gardeners to benefit from

verbal instruction, but also from non-verbal information and the communication of

tacit knowledge.

Most participants described learning gardening skills from others with more

experience in face-to-face interactions. P4 told us that when he was starting out

learning to garden, he spent time talking to experienced gardeners and “picking their

brains.” Even now, he enjoyed being in a community gardening setting, because

with all the activity in the space, “I can learn a lot. I feel like it’s made me a

better gardener.” In our participant observations, we saw the importance of face-

to-face sharing to teach embodied and sensory skills. The first author was taught,

33



for example, to measure ripeness using touch by P8 and P2, to find locations for

incisions on the plant, and the safe handling of pruning shears by P7. Another

anecdote that indicates the importance of face-to-face interaction to communicate

knowledge took place when P9 mentioned that it was sometimes, “hard for me to

explain a plant versus a weed... especially if [the people I am teaching] are newer

to gardening.” The weeds P9 and the first author were looking for in that spot in

the garden were from the dicot plant group. When the first author said that he

didn’t know what a dicot plant was, P9 showed how the orientation of veins was a

differentiating feature. While tracing the outline of the leaf veins with her hands,

she explained: “You have this vein here, but then you have these little veins coming

off the sides ... so they’re not parallel.” This information would have been difficult

to communicate without a shared field of view, gestures, and haptic feedback from

the veins: all elements that can be seen as inherent to face-to-face interaction.

Figure 2.5: P9 showing the first author the difference between leaf venation of a
monocot (left) and dicot (right).

As the first author worked in this setting and was taught to notice plants

in different ways by experienced gardeners, he began to develop a competence for

34



recognizing different plants based on sensory information. P2 presented a contrast

between the leaf texture of pumpkin plants, which the researcher found to be “crack-

ling” (excerpt from field notes), and gourds leaves, which P2 explained were “much

softer than the pumpkin.”

2.5.2.2 Continuing To Learn Outside the Garden

The affordances of face-to-face skill sharing were clear in our findings. How-

ever, as has also been found in past work, gardeners also learned by using a range of

digital resources to find information [27]. In our study, this occurred predominantly

via text and images, such as how-to blogs (P1), social media messages (P4), and

YouTube videos (P2, P5, and P8). P4 described sharing an image of a diseased

plant with people at a nursery to identify the disease, and P1 told us that he posted

pictures of potential weeds online so others could identify them. Individuals also

used digital technologies to get information from those they knew: P3 showed us

a picture she had messaged to her friend so that the friend could remind her of a

name of a plant (Figure 2.6).

Participants appeared to take distinct roles in their online engagements, as

either consumers or producers of information. Some participants (P1, P2, P5, and

P8) acknowledged they were more likely to be consumers of information than cre-

ators. Strikingly, these gardeners had self-identified as experienced and taught us

during participant observations. However, they felt less comfortable sharing their

knowledge online than in-person. For example, P1 has been gardening for several

35



Figure 2.6: Mock-up of a mobile screenshot showing how P3 asked for her friend’s
help with identifying a plant.

years but when discussing helping identify plants on a website he frequents, he feels

he “can’t do that. Online, people know a lot more about [that].” On the other hand,

P7, who had been a master gardener, felt “people are going to learn from me instead

of me learning from them.”

2.5.2.3 Learning From What Is Left Behind by Others

Gardeners also learn without the active involvement or even presence of others

by examining the state of others’ gardens and gardening configurations. Past work

has found that gardeners learn from observing what farmers are doing, for example,

by looking at the produce in farmers’ markets to understand what is in season [5]. We

36



found that gardeners can also learn techniques from observing farms. P8 described

experimenting with a plastic sheet on the ground to prevent weed growth after

observing this strategy in farms and searching online to understand the reasoning

for it (See Figure 2.7).

Figure 2.7: P8 used plastic sheeting after observing its usage by farmers.

One participant observed others’ gardens not just to learn from them, but to

gauge how his garden might be faring. P5 said that when on vacation, “I’ll look at

other people’s plants to see, ‘Okay, where are your tomatoes going now? What’s the

story here?’ ... Because, if I’m in [state X] and we got a heat wave [in] both [state

X] and [P5’s home state Y], I’ll kind of know what to expect when I get home.” By

looking at similar plants in states with similar weather conditions, P5 can gauge

the status of his plant remotely. In these examples, we see that it is not only the

37



experienced practitioners such as farmers whose plots can reveal information and

help relative novices make sense of why they do what they do. Novices can also

compare their progress by observing the gardens of their peers.

Learning new skills via observation of other gardens happened within our re-

search team during the course of the study. We observed that P1, P1’s neighbor,

and P3 all had deer antlers or bones lying around in their gardens to provide nour-

ishment to the animals and the soil (Figure 2.8). During the writing of this paper,

the second author found that the first author had added a small 3D-printed set of

antlers to his indoor plant in the office (Figure 2.9). Reflecting on his experience,

the first author saw this action as a novice imitating the experienced gardeners as

a way to feel more connected to the community of gardeners with whom he had

worked. This goal has been described in learning theory research as motivating and

providing meaning to the process of becoming knowledgeably skillful in situated

learning [3].

Figure 2.8: P1 pointing at deer skull (left). P3 holding deer antlers (right).

38



Figure 2.9: Researcher’s 3D printed antler.

2.6 Discussion

Our findings reveal an understanding of sociality in the garden. We find that

sociality exists in co-located and remote interactions with other gardeners, as well as

with the broader community outside the garden. Practitioners’ sociality preferences

are non-static and diverse. The garden space is designed purposefully in response

to these sociality preferences to allow or restrict access to outsiders or other prac-

titioners. Based on these findings, we present the following considerations for HCI

in designing for collaboration in the garden. Specifically, we consider how the way

sociality exists in the garden bears on teaching or learning skills associated with

39



the activity. We also consider social computing approaches in the garden space for

creating a sense of community inclusion. Given that past work urges designers to be

mindful of the interactions between technology and practitioner sensibilities [5, 6],

we also present potential tensions that arise in introducing technology to the garden

in each section.

2.6.1 Teaching and Learning Embodied Skills in the Garden

Promoting lifelong learning is described as one of the “Grand Challenges” for

HCI [59]. Our work highlights skill sharing as a key domain for design in the garden.

Teaching and learning are regular practices of the participants we studied. Skills are

continuously gained and refined, both through direct interaction with others as well

as through observations of others’ gardens. Yet many individuals lack opportunities

for in-person learning, and a lack of gardening knowledge and access to experts

has been linked to the failure of programs intended to foster gardening skills [24].

Below, we describe opportunities to leverage the expertise of experienced gardeners

as a way to support multigenerational interaction and cultural exchange through

observation and practice.

Learning by doing, under the instruction of expert family members and friends,

was an important way that many participants in our study – as well as the first au-

thor as a participant observer – gained initial gardening skills. Though participants

spoke of using platforms such as Facebook to look up gardening questions or sending

a picture to their friends on the phone, these forms of media are not sufficient for

40



learning many of the embodied skills key to gardening. In other settings, research

has examined ways to support embodied learning through skill demonstration, when

an expert and novice are not co-located. Future work can draw on past work on em-

bodied learning to encourage interaction with sensory stimuli that mimic an expert

practitioner, for example, imitating a projected video of an expert [60] or experi-

encing vibrations synchronized to the movement of an artisan using a tool [61].

There are also open opportunities to create remote real-time skill sharing expe-

riences for outdoor activities such as gardening that draw on the telepresence litera-

ture. This includes real-time interaction between distributed groups of practitioners

using, for example, ego-centric feedback [62] and tangible interaction with remote

physical objects [63]. However, when considering harnessing the expert experience

of gardeners, it is important to note that while some experienced practitioners, such

as participant P7, might feel confident that gardeners will learn from them, others,

such as P1, might be less inclined to share their knowledge. Taking the initiative

to share information, rather than primarily being an information consumer, may

depend on whether the practitioner feels that other people in the group “know a

lot more,” as P1 put it. This self-perception of the practitioners’ expertise relative

to the group is an important consideration when encouraging knowledge-sharing

behaviors. Some may be far more willing to share one-to-one than in larger groups.

Our study adds to past work on how gardeners learn from direct interaction

with other gardeners [26]. We find that learning can also take place through making

sense of the traces left by other gardeners, with some gardeners mimicking techniques

that they learned from their observations. The traces that participants focused

41



on often had to do with gaining an awareness of how more experienced gardeners

might approach sustainability. For example, P8’s technique of plastic sheeting was

based on how farmers discouraged weeds without chemicals. P3 and P1’s ideas

of placing antlers in the garden came from an understanding of the necessity to

nourish the soil and other animals. HCI research on gardening is often motivated

by sustainability [4, 5]. So one design opportunity in this direction is to preserve

and share the traces of skilled gardeners’ actions related to sustainable behaviors

and techniques.

2.6.2 Designing for Varying Social Preferences

Participants created spaces that reflected their sociality preferences using dif-

ferent kinds of boundaries as a way to manage interactions with other gardeners and

people outside the garden. Below, we describe the implications of these findings for

how social technologies in the garden might be received.

In considering social technologies in the garden, it is essential to consider

the ways that preferences for sociality are not constant. In our data, gardeners’

preferences varied with changing moods, tasks at hand, and the constraints and

possibilities of a particular gardening space. In accounting for varying sociality pref-

erences in the design of collaborative systems, researchers have explored concepts

like interruptibility and signaling availability primarily in the context of indoor work

settings [64, 65]. As one example, researchers have studied whether interruptibility

can be estimated from whether productivity is affected by someone typing on a

42



keyboard or standing with one or more guests in the vicinity [64]. In designing or

modifying technologies to be context-aware in outdoor, recreational, and educational

settings like in the garden, how might we translate this concept of interruptibility?

Attending to the location of a gardener and the meaning that they assign to differ-

ent locations is a first step in estimating willingness to be approached. Our findings

reveal how some areas in the garden are assigned significance based on the kinds

of activities that take place. For example, a meditation hut would imply leaning

towards solitude, whereas placing chairs together encourages interaction. Further,

different configurations indicate an openness to interacting with different kinds of

audiences, and whether activities take place inside or outside the garden fence has

meaning. P2 arranging chairs to create a more intimate space for people inside the

garden is intended for close communication with loved ones, whereas the benches

outside the fence of P8’s garden invite unknown outsiders to sit and observe. Gar-

deners’ willingness to interact socially or use technology at all might be estimated

from their locations within the garden and the configurations that they create over

time and in the moment.

Even as we provide implications to avoid introducing technology in the garden

due to its intrusiveness for some, others integrate certain types of technology into

the gardening experience [27]. Expanding on past studies that find gardeners using

technology to coordinate with garden members and showcase their ongoing activities

(e.g., sharing photographs on blogs and social media) [4,27], we find that gardeners

also use technology to encourage involvement in the garden. For example, partici-

pants shared volunteering opportunities and recipes online. And, overall, gardeners

43



reported positive interactions sharing garden-related content. They particularly ap-

preciated receiving compliments. We see the sharing of gardening-related content

as one way to support a sense of community and civic engagement. Here, further

research might consider how sociality varies across different activities or types of in-

teractions and how this relates to activities seen as social or purposeful, for learning

or community building, and by different kinds of practitioners.

2.6.3 Negotiating Inclusion and Ownership

Past work has noted that inclusion is a core value of community garden-

ers [4,55]. Our findings also reveal a desire for inclusion which is carefully balanced

with gardeners’ varying preferences for sociality. This was demonstrated through

a physical configuration of gardens. The tension between the access to skills that

experts can provide to novices and how novices can be excluded due to expert “ter-

ritoriality” [54] is evident with the territory applying quite literally to the physical

spaces of gardens. Below we discuss insights from our findings on how technology

might affect the delicate balance between inclusion and ownership.

From our findings, we see opportunities where outsiders could come to interact

in garden spaces to, for example, learn sustainable behaviors or create a sense of

community by complimenting growers. Researchers can examine approaches that

allow audiences to interact physically with gardening sites through location-based

exploration concepts such as geocaching [66] or even citizen science approaches where

people make data about physical spaces such as backyards, local parks, and other

44



environmental observations accessible to the general public (e.g., Phenology Maps

[67], NatureNet [68], and YardMap [69]). These approaches resonate with aims in

the gardening space, such as learning when a specific plant species will bloom [67]

or the environmental impact of personal growing practices [51]. When designing for

public digitally-mediated interaction in the gardening space, it is necessary to think

about how one might encourage the community to respect boundaries established by

the inhabitants of the space. One approach might support gardeners in indicating

that the local community is welcome to interact with certain elements or parts of

the garden space (e.g., P1’s flags for Independence Day) or inviting volunteers for

particular tasks (e.g., P6’s volunteers helping with weeding and harvesting produce).

The metaphors of different kinds of fences and boundaries to promote or restrict

visibility and access can inspire design in this area.

In our study, gardeners expressed a sense of ownership of the gardening space

and the plants that they cultivate by establishing physical boundaries. Our find-

ings also show examples of participants cultivating relations through their activities

inside the gardening space around their native plants and other memory objects.

What does it mean to design for inclusiveness in a living space whose inhabitants

feel responsible for it, and when the objects inside the space hold meaning for the

people or communities close to them? We propose that there are opportunities in

HCI to support gardeners in highlighting and sharing the meaning that different

objects or plants hold for them and their community. A current project that might

be seen as falling in this design space is the Connected Seeds project which attempts

to connect people to their heritage through food by collecting and sharing stories

45



related to locally-grown seeds [52]. Further areas for connection we identify from our

work include the concept of memory objects that reflect the gardener’s relationship

with a loved one, learning techniques for sustainable growing through observation,

and experiencing different cultures (e.g., sharing produce native to the gardener’s

country).

An important aspect of inclusion is ensuring access for people with disabilities

and mobility constraints – a priority mentioned by gardeners in our study. A fruitful

future direction is to investigate technology’s role in supporting accessible garden-

ing. Research has in the past explored approaches to bringing the experience of a

remote location through an on-site physical proxy to a user. For example, the Tel-

egarden uses a robotic arm to interact with a remote shared garden such as in [70],

and the Teletourism system provides accessible tourism experiences through a video

chat with a video-sharer at the actual physical location that the viewer would like

to experience [71]. This approach of virtually visiting a space might not be ap-

pealing when trying to communicate tacit knowledge that requires certain sensory

stimuli (e.g., learning to determine the ripeness of produce via touch). Further, we

propose that in addition to focusing on enjoyment, engagement, or immersion, as

these prior systems do, it is important to think about how design in this space can

position people and the kind of connections it can enable. For example, volunteers

of different expertise levels connected with experienced gardeners P6, P7, and P9 in

a community learning garden, where they worked with and learned from each other

while contributing to the community’s food security by donating produce. In other

words, the garden is not just a space for recreation and connecting with nature –

46



it is also a meeting point for practitioners that provides opportunities to be good

citizens. One area for future research might involve supporting experts who are no

longer able to garden to remotely share their valuable skills with novices. This kind

of approach could provide access to untapped expertise, a lack of which has posed

challenges to previous projects [24]. In proposing this idea, we do not intend to

minimize the real need to assess and improve the accessibility of outdoor spaces, a

topic addressed in past work through crowdsourcing [72].

2.7 Conclusion

This paper contributes an understanding of how sociality is configured in the

garden environment, and the ways that skill sharing takes place. Through our

participant observations, we found that gardeners use and modify the boundaries

of their gardens to maintain a balance between two considerations: the need for

personal space for themselves, their co-gardeners, or loved ones, and a motivation to

create an inclusive space in-and-around the garden and show community sentiment.

Social skill sharing occurs between gardeners with a focus on on-site interactions

that lead to learning. In addition to direct learning interactions, indirect learning

takes place via observation of other people’s gardens. We contribute a discussion

of design considerations to support interactions for skill sharing between users with

different expertise levels to support varying sociality preferences and in negotiating

the tensions between community inclusion and ownership in the garden nature-space.

47



2.8 Acknowledgements

We thank the members of the University of Maryland and its Extension Master

Gardener Program for their participation in this project.

48



Chapter 3: Probing the Potential of Extended Reality to Connect

Experts and Novices in the Garden

3.1 Overview

In Study 1, I conducted participant observations with experienced garden-

ers to understand sociality in gardening and identify design considerations for skill

sharing in gardening using collaborative technologies. The findings of this study

motivated further examination of remote guidance approaches to address the need

for access to expertise in instructional programs. For these approaches, the study

also highlighted the importance of the embodied sensory experience of a physical

garden, facilitating inspection of other’s plots, and different social contexts involving

family, the local community, and outsiders. Approaches using conventional video

communication can be adequate for visual inspection. However, past HCI research

has noted limitations in remote guidance involving physically manipulating objects

and when conveying the desired embodied experience compared to more immersive

approaches to guidance involving extended reality (XR).

In Study 2 of my dissertation, I examined the potential and limitations of

XR to connect experts and novices for remote skill sharing in skilled hobby ac-

49



tivities like gardening. As XR systems become increasingly available, XR-based

remote instruction is being adopted for diverse purposes in professional settings

such as surgery and field servicing. Hobbyists may similarly benefit from remote

skill sharing. However, little is known about how XR technologies might support

expert-novice collaboration for skilled hobby activities and how they might even be

perceived for an informal hobby space. I had two objectives for my second study:

• Understanding practitioner perceptions of XR for remote skill sharing in the

garden

• Identifying the types of interactions that can be supported in XR for expert-

novice groups.

Study 2 had two parts that involved prototyping and feedback on prototype us-

ability with 27 expert and novice gardeners. From my findings, I discuss design

opportunities and challenges for XR systems in supporting informal connecting in-

teractions and meaningful sensory interactions with a remote environment during

skill sharing. This discussion also raises questions about how the remote environ-

ments (spaces and objects) are meaningfully represented in XR which provided the

primary research direction for my subsequent Study 3.

This chapter is adapted from my paper on “Probing the Potential of Extended

Reality to Connect Experts and Novices in the Garden” [9] published in the pro-

ceedings of the 25th ACM Conference On Computer-Supported Cooperative Work

And Social Computing (CSCW ’22). As the first author, I led the study design,

50



data collection, analysis, and written and oral presentation of this work.1

3.2 Introduction

Extended reality (XR) is a technique that alters a person’s perception of their

environment through the addition of interactive computer graphics over their field

of view [73]. It acts as an umbrella term for a continuum of technologies hav-

ing different variations and compositions of real and digital objects in the user’s

view [74] and includes augmented, virtual, and mixed reality (AR/VR/MR). As the

capabilities of XR devices improve, there has been a reinvigorated interest among

CSCW researchers to understand how the affordances of XR can support remote

collaboration between distributed workspaces. One practical application area that

has seen increasing interest within the larger area of XR for remote collaboration

involves augmenting remote professional assistance and training when performing

skilled physical tasks (e.g., field servicing [1,75] and surgery [2,76,77]). Often, these

systems are designed to improve learning outcomes over traditional video for expert-

novice team scenarios, such as remote experts guiding novices in equipment repair

or maintenance processes [78]. Prior work has presented remote expert XR systems

for teaching other physical activities such as musical instruments [60] or movement

training [79,80]. As inquiry in technology research and design expands outside of the

professional workplace [81] into skilled hobbies [7,82–84], there is an opportunity to

understand whether XR can similarly succeed in supporting remote collaboration in

1Co-authored by Dr. Andrew Irlitti and Dr. Amanda Lazar. Dr. Irlitti helped with feedback on
framing the paper in the context of related work (sections 3.2, 3.3). Dr. Lazar provided feedback
and suggestions throughout the project and paper write-up.

51



a skilled hobby setting. In contrast to professional settings, hobbyist settings possess

differences that might affect the design of XR systems. Skilled hobby activities such

as woodworking or needlecraft privilege the joys of production over the value of the

product [85]. Further, hobbyist learning can focus more on learning a way of life

associated with the activity as a form of serious leisure [86], rather than improving

one’s skill with an economic incentive in mind [87]. Other differences which may

affect user needs and the design of XR guidance include social interaction, commu-

nity history, the strictness of adherence to ethical standards, and if the adequacy of

training is evaluated in an institutional manner [88]. HCI and CSCW researchers are

laying the groundwork to understand how XR might support remote guidance in the

skilled hobby setting. Considerations key for designing XR for this purpose are being

uncovered, such as the ways that experts hone their craft [89,90], the importance of

relaying context for learning physical tasks [91], and perspectives on nurturing sens-

ing capabilities and mentor-apprentice relations through technology [92, 93]. This

paper examines how XR systems might fit into expert-novice collaboration for the

skilled hobby of gardening. Past work cautions us about introducing digital tools

into gardening, as they might interrupt a practitioner’s immersion in nature. How-

ever, socio-technological approaches whose objective is augmenting existing learning

interactions may be more acceptable [4, 28, 94]. Gardening is a particularly fruit-

ful case with which to examine technologies to support learning. Informal social

learning is key to gardening, with practitioners learning from others in commu-

nity settings [95], as apprentices [4], and with family and friends [5, 14]. However,

with deskilling in food production due to industrialization and mechanization [21],

52



there can be a lack of local access to gardening knowledge which may challenge

the implementation of gardening education [24]. Past research has called for fur-

ther study on how technology might support education about food production [4].

Conventional video communication can be inadequate when supporting educational

activities where practitioners physically manipulate physical objects [96]. XR sys-

tems have been well studied in research on remote expert instruction for physical

tasks with established sequences of actions (e.g., equipment assembly [97], surgical

procedures [98]). So, there is an opportunity to understand whether XR could be

a suitable medium to deliver remote learning experiences to distributed gardeners

in a hobbyist setting and when XR environments might actually augment informal

learning experiences.

Our research examines the potential of XR technologies for skill sharing in the

case of gardening. Our research seeks to answer the following research questions:

• What are the perceptions of practitioners regarding remote skill sharing in the

garden?

• What interactions could be supported in XR for novice vs expert gardeners?

• To what degree might users benefit from using XR to collaborate in the garden?

To answer these questions, we conducted a two-part study with 27 gardeners.

In Part 1, we used storyboards and experience prototypes to elicit participants’ at-

titudes toward remote gardening and identify the types of interactions important

to teaching and learning in the garden. From Part 1, we identified three types of

expert-novice interactions: instructing, observing, and discussing. For Part 2, we

53



created XR prototypes to support these three interaction types. We invited partici-

pants to use these XR prototypes in expert-novice pairs to further our understanding

of perceptions of XR and how XR interactions can support or fail to support the

key interactions identified in Part 1.

Our paper makes three contributions. First, we provide results from an ex-

ploratory study as to whether and how to design XR for skill sharing in hobby

activities through a case study in the domain of gardening. We find that partici-

pants were open to remote skill sharing, particularly when there was a motivation

such as the distance between practitioners. In terms of how to design XR for skill

sharing, through participants’ usage and reflection on our prototypes, we identify

necessary affordances to support instructing, observing, and discussing in XR. For

example, supporting orientation in the three-dimensional XR garden space and with

the sun’s position was key for observational interactions. Our second contribution

is in identifying a key dimension for XR to support skilled hobby activities – con-

necting interactions – which have been less central in the professional settings where

much of the prior work on XR for expert-novice skill sharing has been done. This

interaction type involves the ways that practitioners connect personally or socially

to the environment and individuals around them. Third, we discuss the merits and

limitations of XR perceived by expert and novice gardeners for skill sharing. We

discuss challenges and opportunities for the practitioners when inferring information

or conveying the effects of their actions.

54



3.3 Related Work

Below, we discuss past work that studies XR for remote collaboration and

instruction. We provide a general overview of perspectives on individual and social

processes facilitating skill acquisition and specifically discuss existing teaching and

learning practices among gardeners to help contextualize our study.

3.3.1 XR for Remote Collaboration and Instruction

Remote coll