From ‘Social’ Media to Collaborative Media: Cooperative Inquiry

for Shoulder-to-Shoulder Youth Video Authorship Technologies

A DOCTORAL THESIS SUBMITTED TO THE FACULTY OF THE UNIVERSITY OF MINNESOTA BY

Sarah MCROBERTS

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

Dr. Svetlana YAROSH

April, 2020 c Sarah MCROBERTS 2020 ii

Acknowledgements

I would like to thank all of my collaborators, mentors, and broader support net- work. Special thanks to Irene Ye Yuan, Jasmine Jones, and the Harrison Ford Ex- perience: Stryker Thompson, Mikhaila Friske, Evan Sebranek, Chen Chen, Abi- gail Bilger. Gratitude and best wishes to my fellow creators from the "Snapchat Etc." class at Sanford Middle School. Thanks to the Sanford Middle School, especially Lily Thiboutot and Zackaria Antar for supporting this bold collabo- ration. Thanks to Maria Gini and my summer workshop participants. Last but not least, thanks to my advisor Lana Yarosh, my committee, and GroupLens Re- search. I’d also like to acknowledge funding from the GAANN PhD Fellowship, the Jacobs Foundation, and the Google PhD Fellowship. iii

Dedicated to John Carlis.

Here’s a joke I wrote for you: What’s the difference between a steak and a thesis?

I can tell when a steak is done. iv

Contents

Acknowledgements ii

1 Introduction1

2 Related Work4

2.1 Mainstream Online Media and Youth...... 4 2.2 Youth Storytelling...... 7 2.3 Participatory Design with Children...... 8 2.4 Age and Social Networking...... 9

3 Analysis of Video Sharing Practices 11

3.1 Media Choices in Social Networking...... 11 3.1.1 Methods...... 12 Participants and Procedure...... 12 Analysis...... 13 3.1.2 Results...... 14 Age of the Sharer...... 14 Interplay between Sharing Platform and Media...... 16 3.1.3 Contribution...... 18 3.2 YouthTube: Youth as creators on global platforms...... 19 v

3.2.1 Methods...... 19 3.2.2 Results...... 20 Types of Youth Videos...... 20 Types of Youth Channels...... 22 3.2.3 Contribution...... 23 3.3 Share First, Save Later: Opportunities in ephemeral, private sharing 24 3.3.1 Introduction to Snapchat Stories...... 24 3.3.2 Methods...... 25 3.3.3 Results...... 27 3.3.4 Contribution...... 29

4 Cooperative Inquiry 30

4.1 Introduction...... 30 4.2 Related Work...... 32 4.2.1 Cooperative Inquiry and Participatory Design with Chil- dren...... 32 4.3 Methods...... 33 4.3.1 Participants and Settings...... 33 Pre-Teen Youth...... 34 Emerging Adult Designers...... 34 Lead Researchers...... 35 4.3.2 Pre-Teen Sessions...... 36 Typical Session...... 36 Activities...... 37 Action Research Approach...... 38 vi

4.3.3 Undergraduate Sessions...... 38 Typical Session...... 39 Activities...... 39 Action Research Approach...... 40 4.3.4 Analysis...... 40 Workshop Video Content Analysis...... 40 Design Development Iterations...... 42 4.3.5 Ethical Considerations...... 42 4.3.6 Limitations...... 43 4.4 Results...... 43 4.4.1 Practices and Roles in Video Creation...... 44 Skits...... 45 Goofing Around...... 45 Documenting...... 46 Surveilling...... 46 Bothering Others...... 47 Disregarding the Camera...... 47 Technical Moments...... 47 4.4.2 Making Social Play a Priority...... 48 4.5 Video Collaboration Opportunities...... 49 4.5.1 ‘Video Chainz’...... 51 4.5.2 ‘Video-Phone Game’...... 51 4.6 Discussion...... 52 4.6.1 Making More Social...... 53 vii

4.6.2 Collaboration and Creativity in Design...... 54 4.6.3 Mainstream Media and Youth’s Voices...... 54 4.7 Conclusion...... 55

5 Prototype Evaluation 57

5.1 Evaluation of Prototypes...... 58 5.2 Study Design...... 59 5.2.1 Developing and Comparing Prototypes...... 59 Piggyback Prototyping...... 59 Implemented Prototypes...... 60 5.2.2 Participants and Setting...... 63 5.2.3 Session Overview...... 64 5.2.4 Data Collection and Analysis...... 64 Prototype Ratings...... 65 Created Videos...... 66 Qualitative Analysis: Created Videos and in-Session Dis- cussion...... 67 5.2.5 Limitations...... 68 5.3 Results...... 70 5.3.1 Videos Created and General Usage...... 70 5.3.2 Interactions in the Videos...... 70 5.3.3 Evaluations from Youth...... 71 5.3.4 Reflecting on the Curated Audience...... 74 5.4 Discussion...... 75 5.4.1 Understanding Collaboration...... 75 viii

5.4.2 The Interconnected Ecosystem of Social Media...... 76 5.4.3 Implications for Design...... 77

6 Conclusion and Future Work 79

6.1 Summary...... 79 6.1.1 Evaluating Common Media...... 79 6.1.2 Participatory Design with Youth...... 80 6.1.3 Designing for Collaboration...... 81 6.2 Future Work...... 81 6.2.1 Mobile Collaborative Creation: Squid...... 81 6.2.2 Implications for Parents, Practitioners, & Policy...... 82 6.2.3 Implications for Industry...... 83 6.3 Closing Statement...... 84

A Novel Prototype Development 85

A.1 Design Values...... 85 A.1.1 Design Overview...... 86 A.2 Front End Features...... 88 A.2.1 Video Recording...... 88 A.2.2 Video Passing...... 88 A.2.3 Video Combining...... 89 A.3 Back End Video Processing and Storage...... 89 A.4 Conclusion...... 90

Bibliography 92 ix

List of Figures

4.1 Selected examples for video categories (from left to right): Skits, Goofing Around, Disregarding the Camera, and Technical Mo- ments...... 44 4.2 Selected screens from the interactive prototype of Video-Phone Game: Group Activity page (including activities status of all the people participated in the game), Groups Home page (allowing users to create new groups for a new telephone game), Select Prompt page (allowing users to select a prompt to start the game), and Describe Video Prompt (prompting users to continue the game by describing/summarizing the video from last participant). Through interactive prototypes like this one, undergraduate designers in training were able to help ...... 50 x

5.1 These three screenshot images reflect three modes of Instagram (the system piggybacked for this evaluation), group members can view short clips created in the app, like this design presented on the left. This image on the left is a screenshot of a clip created dur- ing the evaluation by the "Jellybeans" group, and features a fes- tive dancing hedgehog with jellybeans and labeled to celebrate the group. The middle image displays a sparse profile page to support the experience of an Ephemeral and Curated Audience. The image on the right reflects the robust editing mode for cre- ators. In the right side image a row of icons on the top allow users to add special effects, download the current image, add gif and static ’stickers’ (like the hedgehog on the left), as well as to draw and write text on the image...... 61 5.2 This figure demonstrates the difference in Collaborative Video Creation. For the "Private Room" prototype individual’s are cre- ating videos with and for each other, but their videos are dis- tinctly tied to their own accounts. However, for the One Story prototype everyone creates and shares as one account, and indi- vidual’s clips are presented as one, unified video...... 62 xi

5.3 This figure presents three examples of the videos created during both sessions. On the left is a video recording of a robot pro- grammed earlier in the camp. It’s demonstrating some editing features (a drawn, glowing squiggle, a "yasss" sticker, tagging a fellow group member, and filtered video). In the center, partic- ipants chose to reenact a viral Vine meme [114] (with water in- stead of vodka). On the right the participant recorded a classic selfie, with slight filter and/or angled effect so that her face is too dark to see...... 69 5.4 These screenshots show examples of interactive creation. On the left, a group coordinates a video of their feet jumping in a circle. On the right, two girls participate in a "camera showdown" first observed in the study in Chapter 4...... 72 5.5 In this figure we see the distribution of evaluations from the youth participants. For these votes the negative rating is assigned to 0 and 100% is assigned to the more positive trait (eg. Confusing=0, Well-Structured=100...... 73

A.1 Squid is designed to help people focus on one video at a time in one project at a time. Each project is the specific iteration of videos that was created with the most recent video. We keep track of the previous version with the parentID. While there could be many videos in a project, one video (especially the first one) could be in many projects as a base that is built off of by many growing projects...... 87 xii

A.2 The technical process for combining videos will employ a “fake it ’til you make it” approach to combining videos. Rather than edit- ing videos together, each video project will function as a playlist that cuts between individually stored videos. The videos will be stored in the backend Firestore Storage, with the project metadata stored and processed in the Firestore database...... 91 xiii

List of Tables

5.1 Key Prototype Insights. This table outlines the foundation and implementation of two critical insights investigated in the proto- type design...... 60 5.2 This table breaks down the descriptive statistics for the number of people in a clip and the number of people actively involved in the clip for both prototypes...... 70 5.3 Contingency table showing the distribution of clips that sparked interactions for each prototype. A Kendall’s τ rank correlation of -0.282 (Z=-2.886, p=0.004) shows that significantly more of the clips recorded with the One Story prototype inspired interactions 72 5.4 Kolmogorov-Smirnov Tests that the means for each factor are dif- ferent between One Story and Private Room find that although the means are slightly lower for One Story, the difference is not statistically significant a critical value for K would have been 1.358. 74 1

Chapter 1

Introduction

Social sharing of photos and videos are one of the most significant uses of the Internet [78]. Open online platforms, like YouTube, are a particularly popular medium for youth who are drawn to the low barrier for entry to reach a po- tentially global audience [26]. Yet, much of the work that has been carried out focuses on identifying and filtering potentially inappropriate content [22] and otherwise supporting children as consumers of media (e.g., the YouTube Kids app [50]). Other work considers the experience of children as subjects in on- line video rather than active creators (e.g., [9, 46]). Relatively little research has been done to directly engage with youth as video creators on these mainstream, commercial platforms, despite their growing presence online. Commercial social media platforms and technologies are creating a new venue for children and teenagers to make their voices heard. However, it is important to understand that current platforms were not designed with the needs of youth in mind and may fail to accommodate the needs, motivations, and skills of this unique but important class of content creators. Considering youth as creators of video content foregrounds new challenges and considerations for designers of Chapter 1. Introduction 2 video tools and video sharing technologies. Privately, teenagers share their sto- ries with friends using tools like instant messaging and videochat [18, 40]; how- ever public sharing is also becoming more common. In 2015, a Pew Research report found that teenagers in the US (ages 13–17) are actively and publically sharing on online social networks, such as Instagram and Twitter [62]. The public response to these developments has been mixed. On one hand, there are concerns about online risks and public exposure. For example, recent work has also investigated teenagers’ perspectives on online privacy and safety [105]. Certainly, it is important to consider where/when children and teenagers may need guidance and support to develop resilience to high-risk content [103] and toxic interactions [106]. However, one can also see youth online activities as an opportunity. We know that video creation and storytelling can play an inte- gral part in learning, especially in the presence of a community that encourages active engagement with educational content [100]. We also see examples of self- expression and identity exploration, like when teenagers use social media to “fashion themselves” [65]. And, even though online activities expose children and teenagers to risk, such exposure can actually help them develop strategies and resilience to such risks in the future [103]. This work seeks to understand and support youth as video creators. While this work will not take a stance on whether or not children should be online, it will reflect and accept the reality that many youth already are creating and sharing online. I believe research must recognize & understand youth behavior as it exists, in spite of policy expectations. It is important to understand and Chapter 1. Introduction 3 maximize the potential benefits of video creation technology, while also mini- mizing the risks. In this thesis, I will describe my foundational work investigat- ing youth video creation on commercial platforms, present a novel exploration of design opportunities with a team of intergenerational youth and detail one specific design insight that blossomed from this exploration, the opportunity for collaboration in video creation. Specifically, I investigate the following research questions: RQ1. What do youth share on video platforms? RQ2. What are the opportunities to enhance the benefits of video technol- ogy? RQ3. How does the experience of co-creation effect the video creation pro- cess? I present a thesis that analyzes the ways that youth use video to connect and express themselves and demonstrates the potential of video technology to fos- ter collaborative play. This thesis will address the Research Questions in three parts: Analysis of Video Sharing Practices (RQ1), Cooperative Inquiry (RQ2), and Insight Evaluation (RQ3). 4

Chapter 2

Related Work

In this chapter, we first summarize existing research on video sharing and high- light the potential and dangers popular social media. Second, we build on the significant, motivating work within the HCI communities that has supported youth voices in storytelling. Finally, we reflect on participatory design (co- design) methods which inspired our strategy of multi-generational youth de- sign teams.

2.1 Mainstream Online Media and Youth

Photo and video sharing online is a significant context of Internet use, particu- larly among young people [78]. YouTube was one of the first online communi- ties to foster a culture where broadcast video content was not only consumed, but also created by the users [21, 27]. A diverse set of other design approaches Chapter 2. Related Work 5 have considered how video sharing could be leveraged to support connected- ness, from synchronous videochat [18], to live-streaming events [92], to asyn- chronously sharing short video fragments [24]. Some of these systems are ex- plicitly designed with a young audience in mind (e.g., [92]), while my work shows that others are appropriated by youths for their own self-expression and social connectedness needs (e.g., [69, 108]). Designers have considered a num- ber of possible interventions to support a reciprocal, productive communica- tion practice with media messaging, including requiring a video response to received messages [96] or asking the users to share a video everyday [80]. How- ever, in many of these cases such sharing becomes burdensome or overwhelm- ing for users [80, 96]. YouTube was one of the first communities to provide a platform and culture for users to both consume and create content [21, 26]. YouTube has since been appropriated by many youth, for their own self-expression and needs for social connectedness (e.g., [69, 108]). In fact, many mainstream social media platforms are now providing an opportunity for children and teenagers to share their sto- ries and experiences. Privately, teenagers use instant messaging and videochat tools to connect with their friends [19, 40]; however, there is also more broadcast sharing on public stages. Pew Research reported that teenagers (ages 13–17) are actively and publicly sharing on online social networks, such as Instagram, Twitter, and Vine [62]. And with popular media sharing apps like Snapchat [88] and Musical.ly [1], users are able to create playful content for both direct and broadcast messaging. Interview and observational studies have been able to tease apart the motivations for young adults to share in these ways [13, Chapter 2. Related Work 6

70, 76, 107], yet little research has directly engaged with younger creators, even though we know that some youth are beginning their careers in content creation as early as infancy (e.g., the “#babyselfies” tag) [72]. However, in considering youth creators on open online platforms, it would also be irresponsible to neglect the concerns around risky behavior and inap- propriate content online. A significant body of work has investigated the care- ful work of parents to monitor and curate the content shared online about their children [8], and more recent work has also examined perspectives of online privacy and safety from teenagers themselves [105]. Certainly, in public on- line domains it is important to consider the guidance and support children and teenagers may need to develop resilience to high-risk content [103] and toxic interactions [106]. Yet there is work that frames youth online activities, and even the risks, as opportunities. Storytelling and video creation can play an im- portant part in learning, especially when educational content is supported by a community that encourages active engagement [100]. Youths can directly inter- act with the creators of educational content through YouTube’s community and commenting affordances [12]. Youth can also use social media stories to “fash- ion themselves” and take chances to practice self-expression [38, 65]. And, even though children and teenagers experience risks online, such exposure can actu- ally help them develop resilience and strategies to manage risks independently in the future [103]. While this thesis does not necessarily advocate for increased or more public sharing by youth, I do choose to respect their goals and potential as media cre- ators. This work aims to meet youth creators at their own levels of aspiration Chapter 2. Related Work 7 and focus on exploring the positive potential of media creation as discussed in the following section.

2.2 Youth Storytelling

While there is little research investigating youth as creators on global, public platforms, there is a substantial body of work investigating and creating op- portunities for children’s storytelling. There are many examples of designs for creative storytelling from creating tangible computing storytelling objects (e.g., [29, 90]), to engaging children in narrative creation for video games [15], to sup- porting them in creating video dialogs with remote peers [48]. Some designs empower youth to use mobile storytelling as a way to create and share their ex- periences with friends and family [16, 37]. Children even broadcast their stories through tools like Canadian Broadcasting Corporation’s StoryBuilder [10]. Across other parts of HCI we have also seen youth video storytelling lever- aged for educational purposes like connecting with other classrooms globally [75], language learning [47], and learning in general early childhood education [112]. A diverse set of design approaches also considered how new video shar- ing systems could support connectedness. There have been a variety of ap- proaches in these video connection efforts including synchronous video chat [20], live-streaming events [92], asynchronously sharing short video fragments [24]. Designers have considered a variety of possible interventions to support a reciprocal and productive communication practice with media messaging, in- cluding requiring a video response to received messages [96] or asking the users Chapter 2. Related Work 8 to share a video everyday [81]. However, these efforts have not yet succeeded in engaging youth creators as broadly and effectively mainstream commercial me- dia. If we, as a community, want to support children in storytelling, it may behoove us to meet them where they are and use their existing practices as a starting point for supporting better storytelling experiences. This is one of my fundamental motivations for exploring youth video creation and sharing through participatory design.

2.3 Participatory Design with Children

The participatory design, co-design, and cooperative inquiry approaches play an important role in inspiring and driving design in the IDC community [32, 41]. While these approaches require time, resources, and the skills to work with children, they can lead to exciting outcomes in the development of new tech- nologies and design-centered learning [32]. Since its introduction to the com- munity, the cooperative inquiry method has been used widely throughout the world and continues to evolve and grow for current research needs [41]. Par- ticipatory work with youth has also expanded to consider the incentives and motivations for teens [42] and encourages youth to be leaders for more of the design process [54]. In my work, I build on the principles and approaches of cooperative inquiry, adapting it for the specific needs of our co-designers and incorporating the contributions of youth co-designers in three different groups Chapter 2. Related Work 9

(aspiring video creator pre-teens, undergraduate designers-in-training, and de- sign researchers). I build on previous work that addresses the dynamic expe- riences of participatory design which includes different generations of young creators [98, 111]. I found this arrangement to be a fruitful way to engage all participants according to their skills and interests and I reflect on the specific adaptations made throughout this paper.

2.4 Age and Social Networking

Multiple studies have looked at social networking practices of different age groups. However, in HCI research, little to no work investigates the motiva- tions of social media sharing across ages. Work looking at newer platforms like Snapchat so far has primarily drawn from college students [13, 76, 107]. In fact, much of the work on social networking draws heavily from college age student populations (e.g. [7, 91, 113]). Additionally, work considering alternative ages focuses on one specific age group: teens (e.g. [Snapchat, Jang_Han_Lee_Jia_Shih_2016_Xie_2014]), older adults age 50+ (e.g. [14, 79]), etc. Ultimately, the different research ques- tions, timings, and contexts for this work limits our understanding of the dif- ferent ways that ages share on social networks. In contrast, Pew Research has addressed social platform usage across ages (i.e. [89, 62, 87]). However, this large-scale survey only enumerated rank and popularity of platforms—such as the rise of Snapchat in young adults, and Facebook’s dominance across ages [87]. These works still do not investigate specific motivations and rationale for Chapter 2. Related Work 10 sharing. In this work we preserve a more detailed slice of social network prefer- ences across ages to provide context in the ways that age and generation impact technology usage. Defining and Operationalizing Youth 11

Chapter 3

Analysis of Video Sharing Practices

In this section we explore three different efforts to understand how young peo- ple use commercial platforms. First, I present the results of a questionnaire dis- tributed at the Minnesota State fare to show how age and media choices affect what people share about their experiences. Second, an in-depth content analysis demonstrates the particular sharing behaviors of youth on YouTube and Vine. Third and finally, an interview study explores the impact of ephemeral, private sharing on content creation.

3.1 Media Choices in Social Networking

Event sharing is an important motivation for social media use, however specific practices, platforms, and preferences may vary depending on a person’s age. To gain a richer understanding of these differences and decisions, we investigated event sharing at the Minnesota State Fair— the center of fun for all ages. Chapter 3. Analysis of Video Sharing Practices 12

3.1.1 Methods

Participants and Procedure

The study took place at the 2017 Minnesota State Fair, in the research outreach building. This outreach building was designed to host University research projects open for the participation of the thousands of daily attendees. Attendees of the fair are representative of the population of Minnesota. This setting allowed us to reach a wide array of social media users and ages. Also, this setting includes participants of demographics frequently left out of social media research, such as people from rural settings. This study was held in a quiet booth. In total, 187 people participated in the study. Ages ranged from 13-74 with a mean of 31. Par- ticipants were 70.8% female and 1.1% other. Of the 123 participants who were at least 18 years old, 119 completed high school degree or equivalent, 72 had completed a college degree, and 22 had completed a graduate degree. 90.9 par- ticipants were from Minnesota, and 1.6% from outside the US. Participants used iPads to complete the questionnaire and had access to a Bluetooth keyboard to write open responses. The main data for the study was collected through questions about types of activities at the fair and how they would share about those activities. The content categories were derived from expert knowledge around state fairs to address a variety of things that someone might share about their experience, and it included an open response to understand the specific content participants might share. These categories were: food, performance, activity/game, exhibit/display. For each of these categories, participants were asked about the likelihood of sharing by media type (write a post, take a photo, Chapter 3. Analysis of Video Sharing Practices 13 take a video, ore take a live video), asked what platforms they would share on, and finally we asked participants to explain why they would share in the way they selected. Following these main categories, we asked follow-up questions which addressed 3 categories of sharing: the people they include in pictures and videos, motivations related to content, and motivations related to network affordances.

Analysis

Quantitative Sections. Quantitative results are reported as either yes/no or on a 7-point Likert scale with 1 being “Strongly Disagree” and 7 being “Strongly Agree.” We assume even interval spacing across all response options. Partici- pants had the option to leave answers blank, and questions never assumed that people would share any of these particular subjects. There were questions at the end asking if people had extra thoughts to add about sharing at the state fair and thoughts about the survey, this gave respondents the chance to share about what else they cared about at the fair, and how they used social media. Quantitative data was analyzed with descriptive statistics, as appropriate. Comparisons be- tween ages and media preferences used Spearman ρ correlations, through lin- ear regression, while binomial questions such as platform usage were evaluated with Contingency Tables, and Kendall’s τ rank correlations as appropriate. For these ranked correlations we bucketed age according to developmental stages as defined by work looking at similar population ranges [59, 94]: young adoles- cent, 13-14 years old (N=29); mid adolescent, 15-18 years (N=42); young adult, Chapter 3. Analysis of Video Sharing Practices 14

19-25 years old (N=33), adult, 26-50 years old (N=42), older adult 50-74 years old (N=41). Qualitative Sections. My collaborator and I used qualitative analysis to pro- cess the open-ended responses to ‘Why would you choose to share this [aspect] in this way?’ section for each of the four aspect categories. We used a constant- comparison approach to drive our iterative process. We excluded empty re- sponses and those without relevant content (e.g. responses solely containing “Nope” or “Because.”). The resulting responses were open coded by my col- laborator with a final corpus consisting of 742 open codes from 170 participants (17 participants chose to leave all sections blank). We clustered the 742 open responses using affinity diagramming to produce themes of motivations. Dis- agreements were resolved through discussion between us. I include these re- sults in this paper as they inform and relate to quantitative sections.

3.1.2 Results

We found that sharing patterns differed across ages, but there were some com- mon motivations for sharing. We begin the section with the way that age played a role in where people shared, consider what and why people shared, and finally look at the interplay between platform and media.

Age of the Sharer

We found that age had a strong connection with where people would share about the fair, and some of their sharing behaviors. There is an age divide in platform usage, with young adults bridging both sides. Facebook is the most Chapter 3. Analysis of Video Sharing Practices 15 popular social platform among our participants (69.5% use it), but also had a strong positive rank correlation with age (Kendall’s τ-b=0.455, p <.001 –see Table 1). Snapchat in comparison, was the second most popular social plat- form (57.8%), but it had a negative correlation with age (Kendall’s τ-b = -0.527, p<.001) that was similarly reflected for all fair aspects. Instagram was almost as popular as Snapchat (56.1%), and like with Snapchat, its use correlated with younger ages (Kendall’s τ-b= -0.396, p<0.001). However, for young and mid adolescents, Snapchat presented as the most popular tool for sharing events with 55% and 52% stating that they would use it across event aspects (com- pared to 20% and 16% for Instagram), respectively. Qualitatively, we notice this reveals a particular platform split with the age divide as many respondents (34) referenced primary platforms— “I usually Snapchat things” (F, 15) and “I use Facebook the most” (F, 59). As people use social platforms to share about events, it appears that older people are more likely to only or primarily use Face- book, while younger people are more likely to primarily use Snapchat for event sharing. However, there is still some as we consider the contingency tables of the bucketed ages with the social platforms they use, we find that the young adult category has Snapchat, Instagram, and Facebook accounts. We see that age plays a role in the social platforms that participants use, and young adults preferences blend between adult and adolescents’. We also see that age plays a role in the experiences and media that people share. For writing posts (see Table 2), the overall mean was neutral for all aspect categories (4.214 mean on a 1-7 scale). However, the older a person is, the more likely they are to write a post about the fair (Spearman’s ρ=0.212, p=0.004). Also, Chapter 3. Analysis of Video Sharing Practices 16 videos were more likely to be shared by younger participants (Spearman’s ρ= -0.186, p=0.011), especially for a performance (Spearman’s ρ=-0.159, p=0.035). Much like the platform preferences, were see there is a contrast between prefer- ences for writing posts and sharing videos. Finally, we also noticed some interesting explanations about sharing what is easiest. We asked participants in the survey to rate how important it was in social networking to share what is easiest. We found that the older the per- son was, the more likely they were to consider ease important (Spearman’s ρ= 0.341, p<0.001). However, when we considered the qualitative responses that referenced ease of use, we actually found more justifications from young peo- ple explaining that their preferences are “because it’s easier to share” (M, 17). In fact, 32 out of 40 relevant comments were from participants under 30. This disconnect could be a sign that people think about ease-of-use differently at dif- ferent ages.

Interplay between Sharing Platform and Media

We expected that platform drives medium choice, but found that medium and aspect can drive platform choice. Feedback revealed photographs to be an im- portant part of choosing the right platform. One participant explained her pref- erence for Instagram with “Instagram is better for pictures,” (F, 21). Addition- ally, another participant preferred Twitter to share about food to avoid sharing a picture, “I don’t have to post a picture with it” (F, 14). Additionally, pictures seem to transition across platforms like Facebook, Snapchat, and Instagram—“I Chapter 3. Analysis of Video Sharing Practices 17 don’t do much live video and pictures get the job done across all three plat- forms,” F, 29 As we dig further into the qualitative codes to understand how people choose different platforms, one participant broke it down precisely: “Instagram and Snapchat are easy to share pictures on while Twitter allows you to communicate with the performers and Facebook allows me to share the event with more of my family members,” (F, 20). Audience is also an important part of deciding when and where to share. In our qualitative coding, we found social motivations to be an important part of sharing. Participants shared their experience “to stay connected with friends and family” (M, 59). Using friends and family as a motivating reason to share was very common with 96 responses coded as such. We also found that audience related to both the collection of people that happen to be on a social network (e.g, choosing Twitter because “old people aren’t on it” (F, 21)) and the volume of people the content could reach (i.e. viral vs private sharing). Some people wanted a wide public audience (“It would go viral.” (M, 16) recording of a performance), while others wanted to be selective of who saw their content (“Instagram is a little more public and Snapchat is a little more private” (F,13) to share on Snapchat). What’s more, some people had specific people in mind for “friends and family,” (43 codes), “a video for extended family to see,” (M, 44) or one participant choose Snapchat because that was where he was friends with the teammates that he wanted to show a robotics exhibit to, and another participant agreed, choosing Snapchat because, “It’s easier to send the picture to certain people,” (M, 17). Sharing with a specific audience can be a designed feature of a platform (like Snapchat), or a more implied part of the sharing experience. It appears deciding between Chapter 3. Analysis of Video Sharing Practices 18 platforms involves a blend of features, audience, and usability. In the following section we discuss how this and other findings can inspire future designs and research. We see that while some platforms stand apart according to the audience they reach, the ease of sharing pictures is still and important motivation for using Snapchat and Instagram. Looking more into this ease of sharing, Snapchat par- ticipants shared that Snapchat allowed for faster, lower quality sharing espe- cially for content like the food pictures that participants mentioned before— “It’s not important enough to warrant a full post, but it’s fun to brag about get- ting it to friends.” (M, 19) and “Immediate, doesn’t have to be good quality” (F, 18). The ephemerality of Snapchat especially seemed appropriate for smaller aspects of a person’s experience, “It’s something small so I’d put it on a tempo- rary form of social media” (F, 16) on choosing Snapchat to post about watching a dog show at the fair. Stories on Snapchat and Instagram were mentioned by 9 participants discussing how they would share so their content would go away in 24 hours. We see ephemerality helps encourage sharing media that might be relevant to others but not important or polished content.

3.1.3 Contribution

This work helped distinguish what young people share versus older adults, and revealed the media choices that people make as they share about their experi- ences. We see that age, content, and audience impact whether someone chooses to share on Facebook or Instagram with a video or a written description. Young Chapter 3. Analysis of Video Sharing Practices 19 people (25 y/o and younger) are especially likely to navigate between an ecosys- tem of social platforms. My design insight takes advantage of this ecosystem for young people to consider a specific experience of media creation, rather than depending on a single platform that is capable of all kinds of sharing (i.e. Face- book).

3.2 YouthTube: Youth as creators on global platforms

Online user-generated video sharing communities, such as YouTube, are becom- ing more popular than conventional studio-produced content. These commu- nities provide every user with the opportunity to create and promote their own video content—a compelling venue for children and teenagers to share their stories and voices.

3.2.1 Methods

To identify the relevant data set of videos, we relied on a crowdsourcing ap- proach to filter videos based on the apparent age of the authors. We began by gathering 2000 recently posted videos using the YouTube API. We asked two Mechanical Turkers (following best practices in [33]) to view each video and describe the author’s age based on available data (e.g., voice, appearance, de- scription, or profile). Videos with no available evidence of age were rated as 0, while videos with evidence were classified based on the involvement level of youths versus adults. Chapter 3. Analysis of Video Sharing Practices 20

We focused on the videos that showed evidence of being created by a youth author (with perhaps minor collaboration with an adult). Additionally, Turkers could mark particular videos as “borderline” in age (e.g., one where it is unclear whether the poster is 19 or older). In these investigations [108, 69] we considered both individual videos, as well as other videos on the same “channel” as the youth authored video that we filtered. A “channel” is a set of videos created by YouTube video authors to publicize and sometimes thematically unite their work (the same author can have multiple channels). By studying individual videos and “channels” we were able to study broad styles of video creation and more personal behaviors and video creation habits. We qualitatively coded the video content and used a constant-comparison, data-driven approach to distill themes of youth video creation. We also used these themes to quantify certain behaviors and styles of videos.

3.2.2 Results

We evaluated the videos created by youth and cautiously compared adult and child video practices.

Types of Youth Videos

We compared the types of videos created to the results of a Pew Internet Re- search questionnaire asking adults about the content of their shared online videos across platforms [77]. Since their study and ours did not use the same method- ology (self-report versus content analysis), we do not make direct comparisons Chapter 3. Analysis of Video Sharing Practices 21 between percentages, but we do examine the relative popularity of each cate- gory. The Pew study found that most users reported sharing the following top four types of videos (in order from most to least reported):

1. Videos of friends and family doing everyday things

2. Videos of themselves or others doing funny things

3. Videos of an event they attended

4. Videos of pets or animals

In contrast to this, the top four most common types of videos found in our youth-authored data set (combining data from both Vine and YouTube), in- cluded:

1. Intentionally staged, scripted, or choreographed videos

2. Videos of friends and family doing everyday things

3. Videos of themselves or others doing funny things

4. Video selfies and expressing opinions

Comparing the Pew data set of adult video authorship [77] with our own inves- tigation of youth authorship, it could be hypothesized that adults view online video as an archive to collect and keep precious memories of everyday life with their family, friends, and pets, humorous moments, and special events. In con- trast, children and teenagers treat online video as a stage. In our data set, we saw them using online platforms mainly to perform (dancing, singing, skits), Chapter 3. Analysis of Video Sharing Practices 22 tell stories (whether capturing their every-day life or staged), and express their opinions and identities in a performative way.

Types of Youth Channels

In the follow up work, we examined the audience engagement practices of young YouTubers (<20 years old) as well as the most popular Adult Youtubers (>20,000 subscribers). We found that many of them (51%) had an awareness of their real or imaginary audience and engaged with the audience through a number of practices, which included acknowledging the viewers, building com- munity by suggesting opportunities for continued engagement, and even di- rectly interacting with viewers who commented on their video creations. With a median of 21 subscribers for YouTube channels, one would think that per- haps many of the youth know their fan base in person (e.g., family and friends). However, none of the youth authors were recorded to be directing their video at this kind of a known private audience. Instead, most greetings and acknowl- edgements were aimed at the YouTube community as a whole. This indicates that most youth were intending to participate in the greater YouTube commu- nity. This is an interesting insight, because most previous storytelling research at IDC and HCI (e.g,. [10, 15, 90]) has focused on the story creation process. Youths also employed many of the same viewer engagement practices as professional quality YouTubers (albeit at a slightly smaller proportion). For example, many children ended videos with requests for likes/subscriptions, pointers to their other online presence (e.g., Twitter), and even collaborations with and plugs for Chapter 3. Analysis of Video Sharing Practices 23 other creators. This is a fascinating context for social learning and future inves- tigations may find interesting fundamental insights on how youths learn and appropriate behaviors as part of larger communities of practice. We compare the youth creation work to channels that they would be likely to see and want to emulate (similar to encouraging child athletes who aspire to- wards professional performance levels). We saw evidence of such emulation in the branded phrases, viewer engagement, and audience involvement practices employed by youth video authors. However, they were not able to emulate other behaviors. The primary difference between pros and youth videos is that youth generally show less evidence of sophisticated editing to increase the in- terest/pace/etc. of the videos. This can both be a reflection of their lack of resources or a lack of experience (since the median number of videos that youth authors had produced was far fewer (27) than the median number produced by professional (832) and all adult YouTubers (170)). However, there is evidence that children and teenagers are interested in emulating sophisticated editing practices. For example, youth authors were more likely to post and edit videos with the aid of freely available templates and tools. This provides some inter- esting implications for the design of technologies and interventions that could assist youths in this process, and we highlight a few of these below.

3.2.3 Contribution

Preforming a thorough content analysis of videos on YouTube and Vine revealed the particular styles of videos that are especially common for young creators. Understanding the efforts that youth make to engage with a public audience, Chapter 3. Analysis of Video Sharing Practices 24 it highlights the importance of recognizing young people as content creators. This directly addresses RQ1 and provides a foundation for future explorations. Considering RQ2, this work shows the potential for remix technology and lever- aging performance or narrative styles of video creation.

3.3 Share First, Save Later: Opportunities in ephemeral,

private sharing

As the third most popular social network among millennials, Snapchat is well known for its picture and video messaging system that deletes content after it is viewed. How-ever, the Stories feature of Snapchat offers a different perspective of ephemeral content sharing, with pictures and videos that are available for friends to watch an unlimited number of times for 24 hours. We conducted an in-depth qualitative investigation by interviewing 18 participants and review- ing 14 days of their Stories posts.

3.3.1 Introduction to Snapchat Stories

Since this feature plays a crucial role in this and the following chapters, I will briefly introduce Snapchat and explain the difference between the Chat and Sto- ries features. When a picture or video is created in Snapchat it can be shared in a variety of ways. Snapchat allows for both direct and broad-cast content sharing, which are then accessed through “Chat” and “Stories” windows, respectively. Chat allows users to share snaps with friends on the app for a maximum of 10 seconds. The Chat feature also allows for text chat conversations, where text Chapter 3. Analysis of Video Sharing Practices 25 disappears after both users leave the conversation, and voice and video calling. This work will reference Chat features only as they relate to the Stories feature or a holistic Snapchat experience, since the Chat feature has been addressed in previous research (e.g. [13, 76, 82, 95]. The Stories feature allows users to create and to view snaps in a temporal sequence (aka “Stories”). Stories may be posted by company sponsors, from thematic Snapchat-organized events, and by other Snapchat users. Apart from company designed Stories, each users Story snap is collected in one unified se- quence, attributed to the user, functioning as an ephemeral timeline. The snaps on these Stories can be viewed any number of times within 24 hours of when they were posted. During this 24-hour period, the sharer also has the opportu- nity to download any individual snap or the entire Story pieced together as a video. Although on Snapchat one can never see how his or her Snapchat friends are interacting with others (such as seeing a friend “like” another friend’s post on Facebook), the app offers a deeper view of how friends interact with one’s own Stories by indicating to the author when any friend has viewed or captured a screen image of a snap. This relative opacity of Snapchat plays a nuanced role in the experience of Stories, which is discussed further for this study.

3.3.2 Methods

We conducted in-depth interviews and followed the participant Stories for 14 days. We did a data-driven inductive analysis to analyze the interviews and snaps. Chapter 3. Analysis of Video Sharing Practices 26

In order to get a specific understanding of what participants shared on their Stories, we created a lab Snapchat account to follow all of our participants’ Snapchat accounts for 14 days prior to the interview. We took a screenshot of every Story snap and every snap sent directly to our lab account. While we ex- pected the study to influence the number of snaps that people shared, we did not want participants to feel pressured to post Stories, so we allowed partici- pants who regularly sent group Chat snaps to also send us those snaps as Chat. During this 14-day period, our participants shared 330 Snaps (M = 17, SD = 20). Our participants were 50% female, ages 18-31 (M = 22, SD = 3). All participants had Snapchat installed for at least one year, and used the app daily. Most partic- ipants considered the snaps recorded in their “following portion” to be typical of what they normally share. The recorded pictures were referenced frequently throughout the interviews by the participants as examples of their typical usage and preferences. After this two-week “following portion” we met the participants for an hour long, in-person interview. We had participants fill out a short questionnaire about their demographics and social media usage. Interviews lasted an average of 55 minutes (SD = 9). In the interviews we asked about a range of Snapchat re- lated questions and reviewed the pictures from the “following period” to reflect on the snaps. We used a constant-comparison, data-driven approach to inductively ana- lyze our interviews and recorded pictures (inspired by the Grounded Theory Method, as described in [71]). All interviews were audio recorded and tran- scribed by the authors. These transcripts were then read and open coded by Chapter 3. Analysis of Video Sharing Practices 27 the lead author resulting in over 900 open codes from the 18 interviews. Two authors then analyzed these open codes, through a constant-comparison affin- ity mapping approach. We open coded the snaps and qualitatively analyzed them in the same way. This paper will discuss themes from the interviews and content analysis in the Results.

3.3.3 Results

Snapchat allows users to create pictures and videos before considering with whom the content will be shared. Stories takes this emphasis on creation first a step further as it offers users the chance to reflect on what they shared over the 24 hours and determine if what they shared is something they would like to save or post again. As P5 explained, the ephemerality of Stories has “an arcade mentality to it,” since posts go away after 24 hours. If he is not proud of a par- ticular story, then he has a fresh slate to try again the next day. P5 also explains that this has changed how he thinks about recording content.

I instantly think, ‘I’m gonna post this,’ and if it’s decent I think, ‘I’m gonna save it.’ It’s a reversal right? You know traditionally you would take a picture and you save it and you would evaluate that and think, ‘Is this worth sharing?’ but with Snapchat, if your go-to is, ‘Let’s snap that and then see if it’s worth saving,’ well and you can go back to your Story and download it and you have this 24 hour buffer, this 24 buffer of content you thought was worth at least capturing and maybe even saving. (P5) Chapter 3. Analysis of Video Sharing Practices 28

P1, P9, and P18 also said that they enjoy reflecting on their recent Stories, and potentially saving them. All participants said that they at least on occasion save snaps to potentially use on other social media. Despite this low-risk way of sharing, users still cared about the way their content was received by the audience. All participants who posted to Stories stated that they appreciated that they could see who looked at each post (and this was information they actively sought out). This gave a chance for reflec- tion on the content. But since Snapchat does not prove a way for friends to respond to specific snaps (e.g., “likes”), participants described a diverse set of strategies for deciding whether participant content was successful enough to warrant saving. P18 judged this success by whether Stories were viewed by one specific audience member—her crush. P9 measured the success of her snaps by whether any of her friends took a screen shot of the con-tent to save for later. Four users specifically mentioned returning to old Stories to see who has viewed them since. These workaround solutions allowed participants to get a pulse on the otherwise self-selecting invisible audience to reflect which Stories warranted archiving for the future. The core mechanic of Stories is that they can be posted and reflected later over the 24 hours, and privately evaluated as a user chooses. They do not have to be created with the intention of archiving every one, nor with the intention of achieving any specific approval from peers (as there were very few ways of measuring this approval). Chapter 3. Analysis of Video Sharing Practices 29

3.3.4 Contribution

This work explores the potential of lightweight, private media creation. The interviews revealed that taking the pressure of permanence away from social media leaves these creators with a sense of freedom. When considering the op- portunities of new media (RQ2), Snapchat demonstrates the potential of social media to incorporate novel features (like ephemerality) into a more playful ex- perience. 30

Chapter 4

Cooperative Inquiry

4.1 Introduction

Social sharing of photos and videos are one of the most significant uses of the Internet [78]. Open online platforms, like YouTube, are a particularly popular medium for youth who are drawn to the low barrier for entry to reach a poten- tially global audience [26]. Yet, relatively little research has been done to en- gage with youth as video authors on these mainstream, commercial platforms. Much of the work that has been carried out focuses on identifying and filter- ing potentially inappropriate content [22] and otherwise supporting children as consumers of media (e.g., the YouTube Kids app [50]). Other work considers the experience of children as subjects in online video rather than active creators (e.g., [9, 46]). It is only recently that our community has begun to investigate youth as cre- ators of video content. For example, investigations of YouTube and Vine reveal that youth use these platforms to tell stories, share opinions, and stage perfor- mances in videos, all while actively seeking to engage with a global community Chapter 4. Cooperative Inquiry 31 of viewers enabled by these platforms [69]. However, it is important to under- stand that current platforms were not designed with the needs of youth in mind and may fail to accommodate the needs, motivations, and skills of this unique but important class of content creators. Considering youth as creators of video content foregrounds new challenges and considerations for designers of video tools and video sharing technologies. IDC’s approaches and methods are partic- ularly appropriate to address these challenges by engaging directly with youth who are currently or aspire to become video creators. In this work, we lever- age cooperative inquiry and co-creation activities within an intergenerational design team to address the following research questions: RQ1. What are the practices and roles of youth as video creators? RQ2. How do current video creation and sharing technologies influence, support, and/or hinder youth as video creators? RQ3. What are opportunities for future technologies to accommodate val- ues, priorities, and experiences of youth video creators? We approached these questions by conducting a 30-week series of work- shops with an intergenerational design collaboration between three groups: 18 pre-teen content creators, 4 undergraduate designers-in-training, and 3 adult investigators. We seek to understand children’s practice of video creating and sharing and their experiences when working with currently available video technologies. We performed qualitative analysis on the data generated from the workshops. We conclude with design implications for building novel video creation and sharing technologies and a discussion of implications for future research. Chapter 4. Cooperative Inquiry 32

4.2 Related Work

In this section, I reflect on cooperative inquiry (and participatory design, co- design) methods which inspired the strategy of multi-generational youth design teams.

4.2.1 Cooperative Inquiry and Participatory Design with Chil-

dren

The participatory design, co-design, and cooperative inquiry approaches play an important role in inspiring and driving design in the IDC community [32, 41]. While these approaches require time, resources, and the skills to work with children, they can lead to exciting outcomes in the development of new tech- nologies and design-centered learning [32]. Since its introduction to the com- munity, the cooperative inquiry method has been used widely throughout the world and continues to evolve and grow for current research needs [41]. Par- ticipatory work with youth has also expanded to consider the incentives and motivations for teens [42] and encourages youth to be leaders for more of the design process [54]. In this study, we build on the principles and approaches of cooperative in- quiry, adapting it for the specific needs of our co-designers and incorporating the contributions of youth co-designers in three different groups (aspiring video creator pre-teens, undergraduate designers-in-training, and design researchers). We build on previous work that addresses the dynamic experiences of partici- patory design which includes different generations of young creators [98, 111]. Chapter 4. Cooperative Inquiry 33

We found this arrangement to be a fruitful way to engage all participants ac- cording to their skills and interests and we reflect on the specific adaptations made throughout this paper.

4.3 Methods

Our methods consisted of weekly afterschool workshops with pre-teen aspiring content creators and weekly design sessions with undergraduate designers-in- training. We describe the roles of all the co-designers, outline typical sessions (and deviations from typical sessions), and discuss our analysis process.

4.3.1 Participants and Settings

Given the far reach of youth video content we actively engaged with an inter- generational collection of youth to recognize their varied approaches to creation and the diverse skills and interests they each bring to bear in the co-design pro- cess. The pre-teen youth met weekly at their middle school as part of an af- terschool program. The undergraduate designers-in-training met weekly at a design space at [their large Midwest US] university. The two lead authors in- corporated Action Research goals [68] in working with both of the two groups to generate ideas and practice design skills. Our motivation for action research was to empower our collaborators to become leaders and designers of the work- shops in a way that would still help us answer our research questions [110]. Given the small size of our groups we are excluding a table of participants, since one could easily use it to attribute each piece of evidence to a specific member. Chapter 4. Cooperative Inquiry 34

Pre-Teen Youth

We collaborated with a local Middle School afterschool program to hold video creation and technology design workshops. We wanted to recruit Snapchat, YouTube, and other online video enthusiasts who would be interested in de- signing and testing new technology. Students at the Middle School had the op- portunity to select this workshop from twelve other activity options. The first author spoke with interested and curious students during the dinner part of the program when all students were together. The school named the workshop "Snapchat, Etc.", and recruiting conversations explained that we were design- ing and creating with all kinds of video creation tools. The lead researchers told students about each day’s activities and the plan to create and inspire new technologies. The school hosted these workshops weekly as an option among other activities. Attendance grew steadily from five to twelve weekly attendees with 18 people attending at least twice (8 males, average of 7 participants per week). At ages 11 to 13 (average 11.6), these collaborators had been sharing on social media accounts for at least one year (average 2 years) and been watching YouTube videos for more years than they were willing to count (average guess 4.5 years). They all had great enthusiasm and experience with video creation. They were roughly split down the middle with their preferences for creating for YouTube, versus mobile video apps like Snapchat, Musical.ly and Instagram.

Emerging Adult Designers

Undergraduate researchers participated in the design process as designers-in- training. Undergraduates were invited to join as collaborators during a lab Chapter 4. Cooperative Inquiry 35

“project pitch” session. The project was described as an opportunity for devel- oping evaluation, design, programming, and research skills. Four designers-in- training (2 women) were involved with the project, all ages 19-24 (average 21.3). On a weekly basis, each designer-in-training participated in design activities and assisted with the analysis of results from pre-teen workshops. These ses- sions were used to develop prototypes and flesh out specific system ideas. The lead researchers presented the ideas and prototypes back to the pre-teens to es- tablish a quick feedback design cycle and further iterate on ideas. As the work- shops continued, 4 of these designers-in-training visited the pre-teen sessions, with 2 joining for the last 16 weeks. By incorporating the designers-in-training voices and experiences, we were able generate more ideas for technology, begin building interfaces and prototypes more quickly in the design process, and cre- ate more iterations of design. Incorporating these designers-in-training created a fortuitous feedback system that provided an amplification of the protagonist voice of the pre-teens to take ideas beyond initial ideation.

Lead Researchers.

In line with recommendations from Yip et al. [111], we find it important to also articulate the roles and participation of the three lead researchers. These re- searchers (including the first two authors) had completed graduate level train- ing in related fields (Computer Science, Communications, HCI, and Design). The first two authors were present for all meetings between both groups. De- spite having advanced skills, all three researchers actively worked to enact their status as equals. Especially with pre-teens, youth were reminded of their own Chapter 4. Cooperative Inquiry 36 expertise as video creators and technology experts. For example, while the lead researchers taught design skills, pre-teens taught the lead researchers about video apps and camera techniques, and undergraduates taught the lead re- searchers about prototyping tools. The lead researchers also made sure to ac- tively participate in ideation and creation activities, offering their own ideas and support. Mutual learning and exploration served as a driving force during these workshops.

4.3.2 Pre-Teen Sessions

The workshop design varied from week to week as the pre-teens developed their interests and established their relationship with the lead researchers. The pre-teen workshops took place once a week (Wednesday) for 34 weeks from September 2017 to May 2018, with a total of 29 sessions (five weeks were can- celed due to school closures). Each session lasted about 70 minutes. We discuss the typical activities in the following sub-sections.

Typical Session

A typical session began with theater warm-up games and then the session’s theme or goal was introduced. In order to most address RQ1 and RQ2, activ- ities around the experience of content creation were common. During a video creation session students were encouraged to create their own video projects in self-selected groups. Participants were provided with different technolo- gies and resources to support their process. Technologies included video cam- eras with tripods (2) and mobile phones (4) with different video apps installed Chapter 4. Cooperative Inquiry 37

(specifically, YouTube, Snapchat, Instagram, Vine Camera, Marco Polo, Musi- cal.ly) and private profiles created for the project. Before participants borrowed the equipment, they told researchers what they wanted to film during that ses- sion. The goal of the research leads was to help support everyone’s contribution and encourage creativity. The last 5 to 15 minutes of each session was spent cleaning up and reflecting on the creations and session in general.

Activities

The thematic arc of the co-design workshops was developed to answer the re- search questions and to help students build skills as designers and creators. The first 6 sessions were spent teaching the youth about design aspects: brain- storming, story-boarding, creating paper prototypes, and evaluating interfaces. These activities were chosen for three reasons. First, they helped offer some context as we answered RQ1. Second, they offered the students a chance to learn more about technology design and help us establish some common de- sign terms for the rest of the semester. Third, they gave us time to get to know the students and stabilize our attendance since it is often a common practice in the afterschool program for students to try several different afterschool activi- ties in the first several weeks. The remaining 23 sessions were spent immersing in the video creation process. Videos were made around the school with session leaders near to support the process. Sessions were also spent in the school’s computer lab to give students the chance to use iMovie and edit some of their larger creations. This emphasis on open exploration gave us a chance to iterate on RQ3, and to help the pre-teens explore RQ2 through experience. Chapter 4. Cooperative Inquiry 38

Action Research Approach

We incorporated an action research approach to developing the specific activ- ities of each session [25, 36]. In answering RQ1, we found it imperative that we take time to foster an equal and open relationship with our collaborators. This ultimately showed itself in the variation on specific activities to arrive at a successful balance of research goals, learning goals, and children’s desires and motivations. For example, during one session of evaluations, students lost fo- cus and started arguing amongst themselves. As we talked about why this is happening, one pre-teen said, “We’re sorry we’re misbehaving. It’s just that we haven’t gotten enough chances to run around.” In response, we made room for everyone to play a running game for the next 20 minutes of the session. This event changed the following sessions, in that we made sure to arrange more ac- tive time in warm-ups and throughout the co-creation activities. It also offered a foundation for students to share their experiences and take ownership of their time.

4.3.3 Undergraduate Sessions

Similar to the workshop sessions with the pre-teens, the activities varied from week to week, to support the interests and skills of undergraduate designers-in- training. The workshops took place once a week (Friday) from September 2017 to May 2018, with a total of 30 sessions (sessions canceled for Thanksgiving, Winter, and Spring breaks). Each session lasted about 40 minutes. Chapter 4. Cooperative Inquiry 39

Typical Session

Unlike the pre-teen creators, the undergraduates regularly continued their own projects outside of meeting times, so workshop times would be spent creating together or presenting knowledge. A typical session began with each partic- ipant reviewing the design or analysis work completed since the last session. The lead researchers also participated in this activity and would use it as a time to give a general update on the activities with the pre-teens. This update would transition into the main activity of session, which was typically a student pre- senting the work that they had done over the previous week, teaching the rest of the team what they had learned.

Activities.

Activities for the undergraduates were designed primarily to echo the activities in the pre-teen workshops and address RQ3. The responses, notes, and other artifacts from the pre-teen sessions each week went to the designers-in-training to encourage them to brainstorm their own ideas and develop wireframe proto- types from the combined ideas. In the first six weeks of the workshops, the ac- tivities were piloting and reflecting on the activities designed for the pre-teens. Members of the undergraduate team had varying experience with design, and were able to take this time to also establish common terms and bring everyone up to speed, as they also took time to explore wireframe and prototyping tools that they would use for developing ideas in the workshop. In the remaining seven weeks of the series, undergraduates used their time to develop designs independently according to their interests. Chapter 4. Cooperative Inquiry 40

Action Research Approach

The undergraduate student group did not require dramatic revisions of group activities, as the pre-teen group. However, we took advantage of varied skillsets and interests of each collaborator to adjust activities according to each person’s individual interests. Since one undergraduate was more interested in app de- velopment, he led investigations into programming tools and tutorials for other students to learn. Another collaborator used design ideas to practice making an interactive prototype. Another used participation in these activities to design their own research questions and join the lead researchers in activity design and analysis for this project. By bringing different skills and goals together, we were able to learn from each other, and invest deeper into the success of both the teams and individuals.

4.3.4 Analysis

We used qualitative analysis of the videos created during the pre-teen work- shops and the iterative design process outlined below to articulate our findings.

Workshop Video Content Analysis.

Primary data for addressing RQ1 and RQ2 in this paper comes from analyz- ing the videos created by the pre-teens. We conducted an in-depth data-driven analysis where each video was reviewed and coded by at least two researchers. Inspired by relevant practices in [69, 108], these videos were coded accord- ing to how the creators engaged with the audience and each other. The first Chapter 4. Cooperative Inquiry 41 author worked with the second and third to complete the open coding process by reviewing the created videos from the first 9 weeks of video creation. The 256 minutes of video resulted in 1000+ open codes, which were then clustered using iterative rounds of constant comparison affinity mapping. These affinity map clusters lead to the development of a codebook with 7 categories of video creator roles and practices. The researchers categorized the previous open codes to refine and apply the codebook. In our final analysis, we tallied the categories that occurred in each video (i.e. if there were 4 open codes of the Surveilling category and 1 open code of Documenting in one video, we marked that the video had Surveilling and Docu- menting). The remaining 314 minutes of video from the later 14 weeks of video creation were then coded by the first three authors and an additional expert qualitative researcher. Each video was assigned to two researchers who individ- ually assigned categories as they reviewed the video. The four researchers cali- brated their coding by meeting and discussing video content for agreement on assigned codes. Borderline and novel interactions that didn’t fit the codebook were noted and discussed among the four researchers. The two sets of codes for each video were then merged, and the two researchers settled disagreements through discussion. The 7 categories of video creator roles and practices from a total of 605 minutes video and 798 coded instances are discussed in the Results as they related to RQ1 and RQ2. . Chapter 4. Cooperative Inquiry 42

Design Development Iterations.

As introduced in previous subsections, we used an iterative design approach be- tween the pre-teen and undergraduate workshops to develop answers to RQ3. During an undergraduate workshop session, we reviewed the artifacts from previous pre-teen sessions (sticky notes from brainstorming sessions, note sheets from interface evaluations, etc.) and began developing a brainstorm list of video app designs. This brainstormed list was organized by themes and then brought the next week to the pre-teen session to collect the responses from the pre-teens. The responses were then brought back to the undergraduates to decide on some of the joint favorite design ideas. These ideas were prototyped in the research lab, and we discuss the ideas in Video Collaboration Opportunities (Section 5).

4.3.5 Ethical Considerations

We took measures to ensure the active engagement & support of all who partic- ipated in the study. Our work was designed to incorporate youth opinions in our activities and adjusted according to the groups’ interests & needs. We ac- tively worked with the afterschool leaders to design enjoyable afterschool ses- sions, keeping answers to our research questions as a byproduct of the valu- able experiences created for the youth. Our study design was submitted to our IRB however it was determined to not be “human subjects” work as partici- pants were collaborators rather than subjects and all activities were judged to be soundly within an educational mission. However, to protect the pre-teen Chapter 4. Cooperative Inquiry 43 collaborators we still collected signed consent/assent forms for pre-teen partic- ipants and received explicit permission from their parents to include the pre- teens as co-authors if they made sufficient intellectual contribution to the work. It is our belief that although they technically did not contribute writing to this manuscript, they were just as much drivers of the activities and investigations as we were. While we wish to respect the privacy of our contributors, it would be inaccurate and unfair to our creative collaborators to not at least include their contributions as acknowledgements.

4.3.6 Limitations

We do not consider the opinions and experiences of our youth creators rep- resentative of all youth creators. Creation experiences were impacted by the devices and technology we were able to provide, the demographics of all of the co-designers, and by holding the Middle School workshops in a school, im- mediately after school. However, we highlight our particular cases of creation and collaboration, to contribute design implications that address some potential opportunities in social media.

4.4 Results

In this section, we discuss the results for RQ1 and RQ2. For these we focus pri- marily on the results of the workshop sessions with the pre-teen youth. Find- ings from the undergraduate designers-in-training workshops will be included Chapter 4. Cooperative Inquiry 44

FIGURE 4.1: Selected examples for video categories (from left to right): Skits, Goofing Around, Disregarding the Camera, and Tech- nical Moments. as points of comparison, but primarily discussed in the following section, Video Collaboration Opportunities.

4.4.1 Practices and Roles in Video Creation

Our analysis of workshop videos revealed that roles in video created depended on the type of video created. We investigated the roles that the pre-teen youth assumed in video creation to have a better idea of the kinds of collaboration that we could design to support. We developed seven categories that differen- tiate the roles and practices in video creation. For example, if a video was a scripted skit, youth worked together to tell a story, with everyone coordinating to do their part. Alternatively, a participant could also use the camera to tease classmates by recording them, in which case the video subjects clearly were not supporting the video. We also noticed that some of these categories came with really specific technical challenges (e.g. editing or wading through useless con- tent). We describe the categories below (with screenshot images of examples in Figure 4.1) and include the main challenges and opportunities of each style of creation. Chapter 4. Cooperative Inquiry 45

Skits.

Videos with skits had scripted scenes and stories that showed a concerted ef- fort by all participants to create the video. While videos were rarely scripted word for word, there were often assigned lines and repeated takes of the scenes that showed the planning involved. The roles of this category were of mutual collaboration between the creators. A repeated example of this was an effort by some of the students to make a news show about the school and afterschool pro- gram. Students filmed segments about school updates, interviews, sports, and weather that they planned to edit together. However, some technical challenges that came with these videos were keeping track of multiple takes and editing separate videos together again. The editing process was much longer than it took to plan and record videos in the first place, although for one of the projects this extra time was because the editor simply enjoyed adding extra flourishes (like sound effects and typed jokes at the end).

Goofing Around.

This category described moments of free play and improvisation between cre- ators. We saw that roles were very relaxed but still collaborative. For example, some students developed goofy characters (like Jet, who lived in a locker, or a Scotsman that lost his leg in the War of 1812), while others recorded themselves hanging out. These videos usually displayed more back and forth banter be- tween video creators than other categories. However, sharing these videos was a challenge — videos made with a lot of Goofing Around were sometimes too goofy to want others to see. For example, at the end of one session pre-teen Chapter 4. Cooperative Inquiry 46 youth picked videos they made to show the rest of the workshop, a group that made an especially goofy video turned away from the screen while it played so they wouldn’t see their own video.

Documenting.

This content focused on the camera holder’s point of view and simply recording their experiences. Creators would pan to different objects or people in front of them and describe them to the camera/audience or they would face the camera at themselves to record their own opinions and reactions. In one session two students set off to explore the hallways and reflect on different rooms, while another filmed a “Behind the Scenes” video of the other’s exploration. Towards the end of the year, 5 of the pre-teens got interested in Documenting with re- action videos of video games or other videos. Like with Skits, the challenge of these videos became combining the separate videos, although with Document- ing there were more instances of creating more side-by-side or picture-in-picture views that were technically difficult, and entirely unsupported by the mobile technology.

Surveilling.

Videos with surveilling were about the experience of filming and the anticipa- tion of content more than they were about recording anything in particular. The primary example of this video is panning around a bustling classroom, when the camera never settles on anything in particular. These videos seemed to be Chapter 4. Cooperative Inquiry 47 more about signaling an interest in video creating than showing the same initia- tive or inspiration of above categories.

Bothering Others.

This category of video recorded a fairly transparent effort to tease other people in the workshop and school, by filming them. While it’s unfortunate that videos were made with unwilling participants, these videos demonstrate the need for more norms around video courtesy and the subtle impact of app designs. For example, a benefit of platforms with shorter video lengths, since a built-in end of the video limits the bothersome interaction.

Disregarding the Camera.

These moments in video creation occurred when the camera was unnecessarily filming or when the recorded video was notably hard to watch. Videos would often be pointed at the ground while creators discussed the next recording, or the camera would be jostled while a creator ran so much that the images were too blurry. While the pre-teens learned to stop and start videos with more in- tention, it could be hard to tell when the camera was shaking too much during the recording process.

Technical Moments.

Instances in this category revealed a wide range of technical details that can go into video creation. These occasions demonstrated when the camera or phone itself dominated the recording, like if creators had to stop what they were doing Chapter 4. Cooperative Inquiry 48 to adjust the camera, or if they were testing a filter. These Technical Moments highlighted the fine line that extra features play in adding to or distracting from the other parts of the creation process. These categories of videos were not mu- tually exclusive. Videos could transition roles with scripted Skits that devolved into Goofing Around or perhaps start with a period of Surveilling before inspi- ration strikes and then start documenting a reflection on a teacher while looking at the door to their room. However, as the type of content created shifted, so would the social dynamics of the video.

4.4.2 Making Social Play a Priority

As the co-designers evaluated current video creation and sharing technology (RQ2), we found the desire and opportunity to make social media more so- cial. Design activities and video creation sessions helped pre-teens develop and demonstrate their priorities in video creation. Our collaborators had strong opinions on their favorite video applications, but a hard time articulating their preferences between these apps. The video filters and effects on Snapchat were held in high regard by the students, but generally students were split between preferring YouTube or the collection of Instagram, Snapchat, and Musical.ly. In- stagram, Snapchat, and Musical.ly differentiate themselves with Snapchat shar- ing ephemeral media that disappears after viewing, Musical.ly supports music and movie clips as the audio for video, and Instagram providing a more general media sharing platform. But these features don’t necessarily succeed in making any one app stand out, as one pre-teen collaborator said with a shrug, “they’re all kind of the same.” Chapter 4. Cooperative Inquiry 49

All of the features in video apps can be a part of fun video creation, but none impacted the workshop videos as much as social interactions did. While groups of creators changed throughout the program, participants quickly lost enthu- siasm to create videos if they weren’t with friends, one day when a pre-teen youth’s best friend was absent, she had a hard time feeling inspired, saying, "It’s just not as fun without [her]." To formalize this point, we drafted a short ques- tionnaire to ask about their priorities around video apps. The questionnaire was completed by 8 participants, and other people (friends, subscribers/followers) were listed as the most important part of video apps. When choosing between “I would rather make videos by myself,” and “I would rather make videos with my friends,” 7 out of 8 participants chose making videos with their friends (though 3 insisted they like both equally), with the other leaving the question unanswered. The 5 undergraduates also completed the questionnaire, where 3 selected the social option (1 insisted on both, 2 preferred making videos alone). So as we designed technology, we were especially inspired by the possibilities of supporting this collaborative experience.

4.5 Video Collaboration Opportunities

We present designs and ideas created by the designers-in-training to address RQ3. The design ideas were developed in cooperation with the pre-teens and undergraduates, with technical implementations lead by the undergraduates. We provide these tool designs as instantiated design implications [83]. Chapter 4. Cooperative Inquiry 50

FIGURE 4.2: Selected screens from the interactive prototype of Video-Phone Game: Group Activity page (including activities sta- tus of all the people participated in the game), Groups Home page (allowing users to create new groups for a new telephone game), Select Prompt page (allowing users to select a prompt to start the game), and Describe Video Prompt (prompting users to continue the game by describing/summarizing the video from last partici- pant). Through interactive prototypes like this one, undergraduate designers in training were able to help . Chapter 4. Cooperative Inquiry 51

4.5.1 ‘Video Chainz’

Inspired in part by the Snapchat Story behavior observed by one of our un- dergraduate designers, Video Chainz is a system of letting people add clips to- gether one after another to create a single chain of videos. In the inspiring video, people on Snapchat would post an image to their story with a prompt to viewers like “screenshot and add a dog” collections of friends would get to see the im- age grow as the picture was shared and modified. We combined this idea with the pattern we noticed starting with the pre-teens’ news crew creating clips of videos from different people that then were to be edited together. To address the challenge of creating complex Skits, we wondered what if there was a system based around the idea many friends contributing? One undergraduate created a design sketch, and the lead authors prototyped this idea leveraging Snapchat on one mobile phone that was passed around as if it had combined videos sent from multiple devices/accounts. The video prototype was presented to the pre- teens and it inspired them to create videos that incorporated more varieties and chains of people, like going around the school and interviewing people they found. This demonstrated collaboration in videos that directly incorporated playfulness and socializing.

4.5.2 ‘Video-Phone Game’

The ’Video-Phone Game’ was partially inspired by the game "Telestrations" [2]. Of the other top design ideas, it was chosen for development as an example of a different kind of ’video game’ (see layout in Figure 4.2), and as an option that was not necessarily about posting the resulting video. The design of the Chapter 4. Cooperative Inquiry 52 game is that one person will film a video of themselves acting out their own simple action or one from a prompt (e.g., ’dancing the jitterbug,’ ’sitting and eating an apple’). Then the video will be sent to the next person playing who will record an audio clip description of what they just watched. Then the next person will record a video and the process of audio and video recordings will repeat until everyone playing has gone. The group then will get to watch the completed video of the process of the action as it changes between people. One designer created an interactive prototype, again the lead authors used Snapchat to prototype the experience. The research lab members that participated in the video enjoyed the game, and encouraged more prototype development. Un- fortunately, the pre-teens were not as excited about the game. As one partici- pant explained, she did not actually enjoy the way things get misinterpreted in games like these, but it did inspire students to think of more games that could be modified to incorporate video creation, like charades.

4.6 Discussion

We propose a model of social video creation, and discuss design implications for making content creation more social on social media. We also consider the ways that our findings in collaborative video echo the collaborative process from this workshop series. Finally, we consider the impact of our collaborators’ ages and the use of social media by youths under 13. Chapter 4. Cooperative Inquiry 53

4.6.1 Making Social Media More Social

Preparing and performing the app critique sessions revealed how similar many video apps are, and it appears these similarities are somewhat intentional [30, 52, 51]. However, videos created in our workshops actually reveal there is room for many different types of social creation and sharing with video. The designs that came from working with the undergraduate designers-in-training also highlighted the possibilities of creating videos across a group of friends, or with video creation as an important part of the game itself. As other research has found, play is an essential role in developing a collaborative community [11]. Games and tangible items can serve as tools for youth storytelling [29, 90]. However the processes and play that these pre-teens developed to collaborate and distribute creation could also be extended to online creative projects in sim- ilar ways. [67]. Working with these youth, we were able to see how formal and spontaneous collaboration contribute to the video creation process. In our workshops, we found that it is more fun to make the videos, than to necessarily show them. Other work has found that different social circles and context col- lapse on social media may also play a part in creators hesitation to share their creations on social platforms [45, 64, 70]. Future work could further investi- gate how playful media development impacts expectations around audience. If media creation itself were more social, perhaps that would change how people choose to share their creations. Chapter 4. Cooperative Inquiry 54

4.6.2 Collaboration and Creativity in Design

Collaborating with both pre-teens and undergraduate designers-in-training helped us bring a wider range of skills to bear. Dividing design activities in a way that supported everybody’s desired and existing skills offered a chance for each team member to provide feedback in a way that felt most valuable for them. Other research on creative collaboration has found that teams of different ages and skills can still collaborate on creative efforts [11]. This work also reflected the importance of taking time to encourage equality and take feedback on all levels. Additionally, it takes compatible interests and status to build a success- ful team [85]. To build this sense of equality we also recommend taking time for reflection [86]. In our study that generally meant taking time to re-watch videos and review design ideas. But we also made sure to incorporate feed- back [109] between groups as a part of the reflection process, sharing and re- sponding to progress from each team. The undergraduate designers might have over invested in the “Video-Phone Game” without bringing it to the pre-teens, but without showing the design opportunities to the pre-teens they might not have expanded their video creation styles (like the video montage of everyone’s smile, and the interview structure for videos).

4.6.3 Mainstream Media and Youth’s Voices

We also want to acknowledge that as people under 13, our pre-teen collabo- rators may not be in strict adherence with their social network terms of ser- vice. While some are using social apps with parent consent and guidance (5 Chapter 4. Cooperative Inquiry 55 participants did report that they access the internet and mobile apps on parent- controlled devices), it is likely that some are not. It is also clear that social ap- plications are available and appealing to youths as young as 8 (the youngest our pre-teen collaborators reported using social media accounts). While there is little academic work to show the full breadth of underage engagement online (Pew neglected to include youth under 13 in their latest account of social me- dia use [62]), the BBC has reported internet access for those ages 12-15 to be ‘on the rise’ [23]. As researchers and technologists, it is important to understand all de facto users of the systems we create, whether or not we are actively inviting them and whether or not their participation is allowed by law. And to that ex- tent, some of the possibilities for these pre-teen users can still be appealing to people of all ages. This multigenerational work opened the door to designing technology that all of us look forward to using.

4.7 Conclusion

To obtain a better understanding of youth content consumers and creators on online video platforms, we conducted a 30-week series of design workshops. We worked with both pre-teen youth and undergraduate students being trained as designers throughout the design process. We qualitatively analyzed field notes and the contents of the workshop videos. From our analysis, we con- tribute an empirical understanding of the challenges faced in the social dynam- ics of the video collection process and empirical evidence that social aspects are the primary reason motivating youth to create and share video. Based on our Chapter 4. Cooperative Inquiry 56

findings, we present challenges and opportunities on how video technologies could be built to support collaborative video creations between youth. Finally, we contribute co-created design instantiations that address these challenges and capitalize on these opportunities. 57

Chapter 5

Prototype Evaluation

The previous studies uncover how features like ephemerality and co-creation bring new energy to the video creation process and lower the barrier for en- try for youth creators. As the final study of my thesis, I present a comparison study of video creation technology designed for a more collaborative video ex- perience. Specifically, I address RQ3: How does the experience of co-creation effect the video creation process? This chapter describes the study design and implementation of investigating the experience of two versions of a video cre- ation prototype, One Story and Private Room. Through this work I show that by comparing the experience of co-creating One Story videos to the experience of simply sharing only with these potential collaborators, I’ve found that although creating one unified video can be disorienting, it does show signs of encourag- ing collaboration among group members. Chapter 5. Prototype Evaluation 58

5.1 Evaluation of Prototypes

In order to understand how co-creation effects video creation, I collected feed- back about a prototype designed to facilitate co-creation. I use co-operative inquiry [41] with a group of young, aspiring technologists to evaluate the way that videos created as one, unified video are different from typical individual video creation and posting experiences. These measures will also contribute to possibilities for future technology. More formally, my research questions for this study are as follows:

1. How does the experience of creating videos differ between the two proto- types?

2. What characterizes videos produced as one collaborative video?

3. What are opportunities for future development of collaborative video tech- nology?

To answer these questions, I designed an afternoon session in which youth used both a co-creation (One Story) and individual creation (Private Room) pro- totype and reflect on their creative process. I report the feedback from the youth and perform a mixed-methods analysis on the videos made. In the videos we note the amount of collaboration and look for themes in the types of videos created. I discuss the typical media patterns and preferences with existing tech- nology and gather feedback about the co-creation experience. Chapter 5. Prototype Evaluation 59

5.2 Study Design

This study took place in an afternoon work shop where participants learned design skills and evaluated piggyback prototype systems for collaborative cre- ation.

5.2.1 Developing and Comparing Prototypes

To investigate the potential of collaborative creation we use piggyback proto- typing as a way to incorporate our design ideas into a realistic and grounded experience for our youth collaborators.

Piggyback Prototyping

This study leverages piggyback prototypes to create a robust experience for youth creators. Piggyback prototyping is a method of rapid social computing prototyping that modifies existing systems to create a novel experience for users [39]. Piggyback prototyping is often used to leverage the critical mass of users on popular platforms such as Twitter or Facebook (e.g. [39, 63]). However, this study chose to leverage the robust recording, editing, and sharing capabilities of Instagram. With the significance of robust editing features, such as filters on Snapchat and Instagram [70,3], we were able to create a more realistic, engag- ing experience for youth to experiment with the prototypes. By using this high fidelity experience, we were able to focus user feedback away from general app usability and feature requests (such as filters, as discussed in Chapter 3) [84, Chapter 5. Prototype Evaluation 60

Critical Insight Ephemeral, Curated Audience Collaborative Video Creation Youth create content with a sensitivity to audience (i.e., YouthTube Youth are drawn to social video Work in Chapter 3, yet there are great creation (i.e., the emergent pattern of Formative benefits to focusing audiences to the remixing from YouthTube work and Connection people already around youth social motivations for using Snapchat (e.g. safety, supporting connections in Chapter 3) beyond social media). Through the participatory design Work on Snapchat (Chapter 3) process (i.e., Chapter 4) we Design suggests ephemeral, curated audiences found potential for designing a Inspiration are an opportunity for fostering a more collaborative structure to support playful approach to creation. these social dynamics. Participants used Instagram accounts (Only for One Story Version) Specific separate from their own, with Participants were logged into the Implementation followers set to only follow fellow same account so that their members of their group. posts would appear as a unified story.

TABLE 5.1: Key Prototype Insights. This table outlines the foun- dation and implementation of two critical insights investigated in the prototype design.

63]. Primarily, we created unique, private Instagram accounts with the added friends and log-ins arranged to replicate specific sharing experiences:

Implemented Prototypes

Although both prototypes, One Story and Private Room, were modified to pro- vide an ephemeral and curated audience, only one prototype implemented col- laborative video creation. We wanted to separate the insight of collaboration, while presenting a novel creation experience for the youth participants to com- pare between. One Story Prototype: this prototype created an experience where each par- ticipant in the group could individually contribute to one unified video. We Chapter 5. Prototype Evaluation 61

FIGURE 5.1: These three screenshot images reflect three modes of Instagram (the system piggybacked for this evaluation), group members can view short clips created in the app, like this design presented on the left. This image on the left is a screenshot of a clip created during the evaluation by the "Jellybeans" group, and features a festive dancing hedgehog with jellybeans and labeled to celebrate the group. The middle image displays a sparse profile page to support the experience of an Ephemeral and Curated Au- dience. The image on the right reflects the robust editing mode for creators. In the right side image a row of icons on the top allow users to add special effects, download the current image, add gif and static ’stickers’ (like the hedgehog on the left), as well as to draw and write text on the image. Chapter 5. Prototype Evaluation 62

FIGURE 5.2: This figure demonstrates the difference in Collabo- rative Video Creation. For the "Private Room" prototype individ- ual’s are creating videos with and for each other, but their videos are distinctly tied to their own accounts. However, for the One Story prototype everyone creates and shares as one account, and individual’s clips are presented as one, unified video. achieved this effect by having each member of the group sign in to the same Instagram account. In the formative work (Chapter 3) we saw that youth were inspired by the social/collaborative aspect of video creation (explained in fur- ther detail in [69, 70, 108]). I was particularly inspired to study Collaborative Video Creation after we found that there was potential for designing a collabo- rative structure in the work described in Chapter 4. A benefit of using Instagram to piggyback our prototype was to ensure that usability aspects of the app (such as video posting and viewing) could be robustly supported. By doing this we would be able to help the participants focus on this grounded experience of creating together. While in Chapter 4 many of the prototyping iterations were low fidelity, or generated example videos, this prototype was able to put the collaborative experience in the participants’ hands. Chapter 5. Prototype Evaluation 63

Private Room Prototype: to compare to the One Story features, we pre- sented the youth with the “Private Room” prototype version. This prototype allowed each group member their own account and created a ’private room’ by only allowing the participants follow each other in the group. This prototype was designed to control against the insight of Collaborative Video Creation in One Story, but still demonstrate an interesting design insight to the participants so that they wouldn’t be biased to favor the One Story version. The partici- pants were automatically following their fellow group members, but discour- aged from following celebrities, friends outside of their current group, or even their own, usual Instagram account (these limitations were also present for the One Story prototype, but not presented as the primary feature as it was for Pri- vate Room). This designated collection of friends represented our specific exam- ple of the Ephemeral, Curated Audience insight. Inspired by the self-selected audience of Snapchat Stories (Chapter 3) and the hard-to-share goofy videos of Chapter 4.

5.2.2 Participants and Setting

My three assistants and I completed the evaluative workshop during an after- noon class at a programming day camp of 24 middle school and high school age girls (two girls chose to use the time to work on their camp projects instead of participate in the study). Although we did not collect demographic information about our participants, this day camp hosted girls entering 6th-11th grade in the coming fall (average and median 8th grade). This camp also recruits a diverse population of attendees (reporting 17% Black/Hispanic). Of our 24 participants, Chapter 5. Prototype Evaluation 64 all but one had prior experience with Instagram, and over half chose to use their own phones for a more authentic and comfortable evaluation of the systems.

5.2.3 Session Overview

The three hour session consisted of an introductory lesson on Human-Computer Interaction that focused on social media technology followed by time spent us- ing and reflecting on the piggyback prototypes. The introduction to HCI and prototyping served as a way to establish a common ground for evaluating. We then split the participants into five groups and had the participants apply what we just discussed and their years of experience on social media to pick the most important factors to consider for the prototypes from the AttrakDif affinity tests [43,4] Also during this time we distributed loaner phones and guided the log- in process. Then participants took time to use their mobiles devices to create videos with both prototypes (first condition, One Story or Private Room, ran- domly assigned to each group, and then the other in the second round). After each session with a prototype, participants used sticky notes on the wall to rate the performance of the prototype on spectra based on the factors most voted for by the participants: “Well-Structured/Confusing,” “Connecting/Isolating,” “Pleasant/Unpleasant.”.

5.2.4 Data Collection and Analysis

The analysis for this study draws from three main sources: rating results (from the AttrakDif factors), videos created, and session recordings. This study uses Chapter 5. Prototype Evaluation 65 quantitative data from the youth participants’ ratings and a mixed-methods analysis of the videos created by the participants.

Prototype Ratings

While we expected that participants would engage with the video creation pro- cess to varying degrees, I wanted to make sure that all participants had the chance to provide feedback about the systems but still maximizing the available time to actually use both prototypes. By having participants get up and put a sticky note on the wall, we provided the participants a chance incorporate some physical embodiment of their ratings [6]. Before the prototype sessions started, we presented the teams with a list of evaluative measures from the AttrakDiff affect measure, and had each team pick the five measures that they thought would be most important for evalu- ating social media. We then tallied the selections from all the teams. As we tabulated the results, three measures received the most votes from 4 (of the 5) groups. These top three measures were announced to the group: “Well- Structured/Confusing,” “Connecting/Isolating,” “Pleasant/Unpleasant.” We taped these printed words up to three separate walls (one for each spectrum) at approximately 6 feet apart. After the first session participants placed a sticky note with a "personal codename" written on it (to obfuscate potential judgement from peers, and so we could track what they were rating), and after the second sesson they moved their sticky note for their second prototype. We recorded the ratings with photos taken head-on for each picture, and then a research as- sistant used pixel analysis to generate a 0-100 number for each rating (0 being Chapter 5. Prototype Evaluation 66 closest to the negative ratings: ‘Unpleseant,’ ‘Confusing,’ ’Isolating’). These rat- ings are presented in aggregate in Figure 5.5. Since the collection of responses are relatively small and not normal, I applied a Kolmogorov-Smirnov test to compare the within subjects ratings for both prototypes for each of the 3 fac- tors. A Kolmogrov-Smirnov test does not assume a normal distribution and is particularly sensitive in the middle of the distribution, so its an effective way to understand the change in ratings from the participants. The results provide a structured, self-reported answer to RQ1 for this chapter: How does the experi- ence of creating videos differ between the two prototypes?

Created Videos

In addition to collecting the self-reported data about the experience of using the prototypes, I also studied the videos created. Although many of these stories maintained or returned to a central theme between the clips, the variation of size of the stories between groups and sessions makes it challenging to mean- ingfully compare entire stories. So, in line with the practice in the Snapchat study in Chapter 3, this work breaks down the whole stories, and analyzes the individual clips that create the stories. I applied mixed-methods analysis to the clips created in this workshop by qualitatively analyzing the content of the videos and recording the ways that the participants interacted with the camera. Inline with Chapter 4, I assigned open codes to the videos collaboratively with a research assistant. To go beyond Chapter 4, I also wanted to measure the ways that these youth Chapter 5. Prototype Evaluation 67 used video creation to start interactions (e.g. prompting people who aren’t oth- erwise paying attention, using the act of filming to start a conversation like with ’bothering others’ in Chapter 4). With a supporting researcher we counted the number of people in each video, the subjects who engaged with the video (e.g. looking at the camera, gesturing or speaking in response to the camera person), and whether or not the video caused someone who wasn’t paying attention at the start of the video to begin interacting with the camera person. A research as- sistant and I individually coded each video and then met to discuss differences in codes and come to a consensus. The counts of the subjects and the people interacting with the camera were non-normal, so to compare the differences I apply a Welch’s t-test. However, whether or not filming sparked an interaction, was binomial and evaluated with Contingency Tables and Kendall’s τ rank cor- relation. These quantitative results provide perspectives for both RQ1 and RQ2 of this chapter.

Qualitative Analysis: Created Videos and in-Session Discussion

In order to get a rich understanding of the experience of using these prototypes we collected and analyzed the videos created by the youth (similar to Chap- ter 4, and the YouthTube research in Chapter 3) and recorded conversations we had with the youths during their sessions with the prototypes. These recorded conversations gave us the chance to collect in the moment feedback from par- ticipants without them needing to take time to write down thoughts, and gave us the chance to observe behind-the-scenes recording dynamics. Chapter 5. Prototype Evaluation 68

Oddly, although we used a camera to record these discussions, participants insisted that we don’t specifically record there faces talking (even though they were featured in the videos created). In the spirit of collaboration we used these recordings as audio, and collected a limited amount of discussion totaling 15 minutes of analyzed discussion. The feedback recorded was transcribed and open coded by an assistant researcher. These open codes from the recordings were incorporated with the 237 codes from the videos created by the partici- pants and analysed through constant comparison affinity mapping as described in previous work (Chapters 3 and 4). This analysis helps us better understand why participants chose their ratings and how they approached the video cre- ation process.

5.2.5 Limitations

While the structure of this evaluation allowed for a close working relationship with our participants to encourage honest and empowered feedback, two ma- jor limitations of this work are the all girl collection of participants and the workshop-style workshop for this comparison. Although publications like Fit- ton and Read’s [35] argue that all girl populations for participatory design activ- ities can be entirely appropriate, I acknowledge this may bias the results. Ad- ditionally, this single workshop limited the amount of time for reflection and experimentation of participatory design compared to Chapter 4. Due to techni- cal limitations, not every participant would be able to have the same experience with the prototypes. Participants with varied phone experiences and occasion- ally stalled WiFi in particular. However, by working to build mutual respect Chapter 5. Prototype Evaluation 69

FIGURE 5.3: This figure presents three examples of the videos cre- ated during both sessions. On the left is a video recording of a robot programmed earlier in the camp. It’s demonstrating some editing features (a drawn, glowing squiggle, a "yasss" sticker, tag- ging a fellow group member, and filtered video). In the center, participants chose to reenact a viral Vine meme [114] (with water instead of vodka). On the right the participant recorded a classic selfie, with slight filter and/or angled effect so that her face is too dark to see. with these youths in our activities, I believe that this work empowers the voices of an otherwise ignored or discouraged population in social media [49, 34], and empirically reports on their experiences. I emphasize that this work is pointing to emerging correlations, not proving causation. I look forward to future work that will continue to explore the insights demonstrated in this work. Chapter 5. Prototype Evaluation 70

Prototype N Mean SD SE One Story 66 1.879 1.75 0.215 # people in clip Private Room 40 1.7 2.09 0.331 # people actively One Story 66 1.167 1.399 0.172 involved in clip Private Room 40 0.6 0.778 0.123

TABLE 5.2: This table breaks down the descriptive statistics for the number of people in a clip and the number of people actively involved in the clip for both prototypes.

5.3 Results

5.3.1 Videos Created and General Usage

There were a total of 106 individual posts collected in these larger arrangements. Of these 106 videos, 66 came from One Story and 40 from the Private Room prototype. Many videos used some kind of filter or editing affect in the post (23/60 for One Story, and 26/40 for Private Room). One group was an outlier with 29/60 clips in a One Story collection (so the whole group contributed to this), and 2 groups only posted one clip to their story for one session. While 13 participants only posted 1 clip in Private Room, due to technical limitations of Instagram we’re unable to determine how much individuals posted for One Story.

5.3.2 Interactions in the Videos

As seen in Table 5.2, although there are roughly the same number of people in each clips for both prototypes, there are slightly more people actively involved in the clips. I also found that more interactions occurred with the One Story prototype than Private Room. Some of these interactions were inviting group Chapter 5. Prototype Evaluation 71 members to try out filters. For example, in one set of example videos, as the camera panned around the girls in the group, their faces changed to be more masculine. This surprising filter inspired shouting reactions like ‘I look like my Dad!’ On the other hand, many other inspired interactions were filming the other group members, similar to the ‘bothering others’ category recognized in Chapter 4. Responses to this ‘bothering’ ranged from pushing the camera away to smiling and striking a pose. Screenshots from these collaborative examples in Figure 5.4. Table 5.3 shows the distribution of posts on One Story and Pri- vate Room that sparked an interaction. A Kendall’s τ rank correlation of -0.282 (Z=-2.886, p=0.004) shows that significantly more of the clips recorded with the One Story prototype inspired interactions (counts broken down in the contin- gency table, Table 5.3. One participant reflected on how the difference of the two prototypes changed the sense of interaction:

Well, the first time [using One Story] we were all together and posting and doing stuff. And then the second time [using Private Room] we were alone and even though we were posting about ourselves or whatever, we were just not, like, together. It didn’t feel that interactive.

5.3.3 Evaluations from Youth

The workshop participants votes revealed 3 top factors: “Well-Structured/Confusing,” “Connecting/Isolating,” “Pleasant/Unpleasant.” Twenty-four (24) participants rated both apps according to these scales, as shown in Figure 5.5. Through our evaluative ratings we found that both prototypes were rated moderately. Across Chapter 5. Prototype Evaluation 72

Camera Camera Did Prototype Sparked Not Spark Total Interaction Interaction One Story 45 21 66 Private Room 37 3 40 Total 82 24 106

TABLE 5.3: Contingency table showing the distribution of clips that sparked interactions for each prototype. A Kendall’s τ rank correlation of -0.282 (Z=-2.886, p=0.004) shows that significantly more of the clips recorded with the One Story prototype inspired interactions

FIGURE 5.4: These screenshots show examples of interactive cre- ation. On the left, a group coordinates a video of their feet jumping in a circle. On the right, two girls participate in a "camera show- down" first observed in the study in Chapter 4. Chapter 5. Prototype Evaluation 73

FIGURE 5.5: In this figure we see the distribution of evaluations from the youth participants. For these votes the negative rating is assigned to 0 and 100% is assigned to the more positive trait (eg. Confusing=0, Well-Structured=100 all categories, means ranged from 42.7 (Well-Structuredness for One Story, std. 20), to 58.3 (Private Room, Pleasantness, std. 19). Kolmogorov-Smirnov distri- bution analysis failed (Table 5.4) to find a significant difference between One Story and Private Room ratings for any of the 3 factors. As the data suggests some participants did not feel like there was a distinct enough difference to change their ratings, ’Well, I thought mine [,the ratings,] were the same.’ While there was not a statistically significant difference between the prototypes in our conversations with the youth participants, we learned that the structure of One Story created some extra confusion for creators. Since there was no way to attribute who created what videos, participants felt like it was harder to personally create a series of videos that could be interrupted, and that the interrupting videos seemed to come from nowhere. ‘It was kind of confus- ing how someone puts something on my story but I didn’t know what it was. I was Chapter 5. Prototype Evaluation 74

Kolmogorov-Smirnov Tests K p Well-Structured 0.737 0.649 Pleasant 0.577 0.893 Connective 0.295 1.000

TABLE 5.4: Kolmogorov-Smirnov Tests that the means for each fac- tor are different between One Story and Private Room find that al- though the means are slightly lower for One Story, the difference is not statistically significant a critical value for K would have been 1.358. kind of confused.’ However, some participants preferred the the One Story proto- type because they could automatically see the videos they made together with one tap, saying ‘Yeah. I mean like, we weren’t sharing one account, so if we really wanted to see what everyone else is posting, we had to, like, go through all the accounts and it wasn’t connected in any way.’ So perhaps some explanation for the similar scores is that the pros and cons of the prototypes canceled each other out. As we noticed the video types and recording behaviors, a possible explanation for these different experiences could be the dependent on the type of video activity that participants engaged in. One group coordinated their videos entirely (e.g. shooting different angles of the same skit), while other groups mostly filmed their own individual ideas while sitting or standing together as a group (e.g. one person asking a friend to ’say something weird’ and then another group member talking about her devotion to Bruno Mars).

5.3.4 Reflecting on the Curated Audience

Privacy emerged as an important part of participants’ experience with both pro- totypes. Each group made sure to ask who would be able to see these videos. Chapter 5. Prototype Evaluation 75

They were generally relieved to learn that the posts were kept entirely private, although one group of participants wanted to follow and be followed by their personal accounts on Instagram. Participants described a sense of freedom be- ing able to create videos that only this group of relevant friends could see.

Cause I have Instagram, like, a separate account, it feels different posting for everyone rather than just people that you’re sitting right next to that you know it’s going to be funny if you post anything. Rather than [using Instagram usually] worried if you’re posting something and what everyone else is going to think about it.

Since this more private sharing was a feature of both accounts, I wonder if the significance of that experience is also part of what contributed to our inability to statistically differentiate between the two means (i.e. it was so meaningful that participants had a private way to share and view posts, that it didn’t matter as much whether the posts were combined into one video or kept separate).

5.4 Discussion

5.4.1 Understanding Collaboration

Although this was a relatively small group of participants, we explored a range of methods to consider when observing collaboration in video creation: pro- totype ratings, behavior in videos, recorded reflections during sessions. This mixed methods approach became a crucial part in understanding the ways that Chapter 5. Prototype Evaluation 76 both prototypes influenced creation. In this work, I believe that collaboration in- fluenced the creation of more videos. For example, ‘camera showdowns’ create two videos out of one moment, but also, another group often created multiple videos of the same moment by filming it from different angles (as discussed in Figure 5.4 and Chapter 4). As other work has found, capturing content like pic- tures can make events more engaging [31]. Future work could explore ways to incorporated those varied angles in a final product [60] or explore if these per- spectives are a means to an end product or if the end is the engagement of the moment.

5.4.2 The Interconnected Ecosystem of Social Media

This study further reinforced the ways that social media applications are inter- connected for young people. Despite the “death” of Vine [97] in 2016 (rest in peace), one of the main creation activities for one group was reenacting a viral vine known as “Two Shots of Vodka” [114] (the joke of the video is that while a cooking instructor announces “two shots of vodka” for a drink, she contin- ues to pour for much longer than two shots would require). When I asked the group why they were laughing about vodka, they asked if I knew what Vine was (rude), and they went on to explain that even though Vine ended when they were as young as 8, treasured Vine videos were still popular and preserved in YouTube compilations. Similar to the remaking/remixing category of video creation (Chapter 3), these participants chose to create some of their videos for the prototype as a remake of a Vine. Chapter 5. Prototype Evaluation 77

Another reason social media apps intermingle is through leveraging differ- ent creation features and audiences. During a discussion a participant in a group casually announced “Snapchat has the best filters.” We saw this demonstrated as the gender changing face filter mentioned earlier, was actually not supported on Instagram. One participant actually recorded videos on her own Snapchat ac- count, saved the video (without posting), and then uploaded it to the One Story prototype. While this transition between applications may have been driven by the structure of the workshop, it at least demonstrates that these creators are willing to switch to an application with the editing features they want than work with a selection that might feel limited. Ultimately, this could be a great opportunity to design for and anticipate new interactions. Future work should do more to explore this ecosystem for content creation. As suggested in this work and Chapter 3, if creators are willing to blend their experiences between applications, perhaps it should be easier to share across multiple services, or we can see how even short lived applications (like Vine) can have a lasting impact on the ways that people share online.

5.4.3 Implications for Design

While this work reveals potential for collaborative technology, it also uncovered some challenges. While I hope this work inspires further design of collaborative technology, future work should do more to actively consider the structures im- plemented for collaborative video. Participants expressed confusion trying to understand what order videos were happening in, and they wanted to be able to know who was contributing Chapter 5. Prototype Evaluation 78 what. While some work [61, 99] suggest that anonymity is supports personal expression online, this work shows that it may not be as helpful in co-located, collaborative creation. Another limitation of the One Story prototype is that it also took away the ability for individuals in the group to respond to each other. Inspired by this and my formative research discussed in Chapter 3 [69], I rec- ommend incorporating more editing and video features to help youth organize and distinguish their content from their phones. For example, I suggest incor- porating screen-names, icons, or other branding on individual clips in a larger video project in order to help assign credit or provide clarity during the film- ing process. Another option could be to allow creators to reorder their videos (especially to help creators trying to create a multi-clip narrative). The critical insight of the Curated Audience was more inspiring to the partic- ipants than I expected. It took some monitoring and extra explanation with the participants to not immediately follow their current accounts, favorite celebri- ties, and friends outside of the workshop, but two groups were especially struck by the freedom that only creating for their immediate audience gave them. Sup- ported by my earlier research on Snapchat (Chapter 3), I encourage more social media services to give people the chance to shake off their usual whole audi- ence. 79

Chapter 6

Conclusion and Future Work

In this chapter I summarise the contributions of the previous chapters and con- sider the broader implications of the work described in this thesis. By focusing on hearing and amplifying youth voices, this work brings an important and novel perspective on social media technology.

6.1 Summary

6.1.1 Evaluating Common Media

This work presents an thorough evaluation of influential social media to an- swer RQ1: What do youth share on video platforms? By exploring the presence of youth on YouTube and Vine and understanding the distinct ephemeral affor- dances of Snapchat Stories, I demonstrated the impacts that platform and media styles have on sharing. Significance of Platforms and Audience: YouTube is a global platform, where we observed youth as performers reaching out to their broader audience. In Chapter 6. Conclusion and Future Work 80 contrast, the way that Snapchat creators thought about their audience (poten- tial audience members will watch if they’re interested, but won’t be bothered if they aren’t) meant that creators were more concerned with expressing their own interests than performing for other people. Significance of Media Styles: While Vine had a similar, global platform as YouTube, its short time limit and looping replay attracted a strong teen pres- ence. Ephemerality shows the potential for more creativity. My cooperative inquiry work expanded this investigation by developing platforms for gami- fied creation and social experiences, beyond even creating a final product. In the survey around event sharing we see that young people are most likely to manage a collection of platforms to suit the variety of content that they might share.

6.1.2 Participatory Design with Youth

The methodology of this work is also a contribution to the field to answer RQ2: What are the opportunities to enhance the benefits of video technology? Through empowering youth voices, we were able to get a new perspective on the poten- tial of social media. By leveraging two different generations of youth we were able to bring about new ideas and designs that would not be possible with only one age group. While youth under 13 years old are not allowed in a platforms terms of service, this does not mean that they aren’t active on those services. Chapter 6. Conclusion and Future Work 81

6.1.3 Designing for Collaboration

The final part of this work reveals the potential of innovating social media be- yond generating media for posting. Specifically, I address RQ3: How does the experience of co-creation effect the video creation process? This chapter de- scribes the study design and implementation of investigating the experiences of two versions of a video creation prototype, One Story and Private Room. Through this work I show that by comparing the experience of co-creating One Story videos to the experience of simply sharing only with these potential col- laborators (Private Room), I’ve found that although creating one unified video can be disorienting, it does show signs of encouraging collaboration among group members.

6.2 Future Work

6.2.1 Mobile Collaborative Creation: Squid

The findings in this thesis are currently driving the development of a collab- orative video creation application– Squid. Squid implements collaboration by supporting video creation among people in the same place, with their own de- vices. Videos are created by one, short segment at a time to encourage collabo- ration and play in the video creation process, similar to the One Story prototype. A key improvement of Squid is that users have more control of how their clip contributes to the combined video. After one user receives a video from a (co- located) friend, they have the chance to record their clip and place it in exactly Chapter 6. Conclusion and Future Work 82 the spot they want it to appear. This addresses some of the confusion that par- ticipants described in Chapter 5. More details about Squid are included in Ap- pendix A. Future work on Squid will be able to investigate this novel structure for collaborative video.

6.2.2 Implications for Parents, Practitioners, & Policy

This thesis reveals some significant gaps in the ways that policy expects youth to exist (or not) online and the ways that they actually are. This disconnect has inspired several moral panics in recent years [17, 93], and lawmakers and platforms have reacted with various measures: increasing penalties for users under 13 and placing half hour limitations on social media for everyone [73, 74], proposing an age limit of 21 to own a cellphone [5], or increasing the required age to 16 [28]. However, I encourage policies more informed by youth voices. I agree with fellow researchers that engaging in social media online can be a useful part of learning about risks [55]. Fundamentally, I believe that completely removing children from risky situations does not better prepare them for those situations in the future. The presence of youth online creates many nuanced challenges for youth and adults from managing the popularity of ’unboxing videos’ [56], to considering the health of child stars on YouTube [101], to managing interpersonal risks [104]. However, these issues also reflect the undeniable presence of social media in the lives of youth. I believe that a more impactful approach to online safety is one grounded in the reality of children’s presences online. To this end, I’ve presented my research to practitioners and members of Society of Research in Chapter 6. Conclusion and Future Work 83

Child Development. I’ve also inspired more (3) after school programs to feature social media groups in order to support more thoughtful and connected social media use. Future work will continue this bridge building and help develop policies that balance protection and empowerment for youth.

6.2.3 Implications for Industry

Using an ethnographically inspired approach with users, I have found that even young, underestimated users can provide inspiring insights and drive innova- tion. I’ve had the chance to present this work to Snapchat Researchers and dis- cuss the potential of more collaborative media. I am also bringing this collab- orative spirit to technology creation itself as I am currently working at Google performing user experience (UX) research on tools created for software devel- opers. Although there are significant differences between software developers and middle school youths (you’d be surprised), by bringing cooperative inquiry methodologies to this technical space, I am able to help technologist engage more with UX research and better understand the value of user-centered de- sign. Although participants and subjects in these studies point to more potential for richer media experiences, more engaging features, and phones more capable for editing and recording, my research in Snapchat and with youth creators also points to craving for simplicity, at least at a social level. There’s a certain free- dom that creators feel when they know that their audience is only those who want to take part [70], or only the people with whom they’re creating (Chapters Chapter 6. Conclusion and Future Work 84

4 and 5). I have no doubt that social media services will find new and innova- tive ways to make video creation engaging (e.g. the sensational experience of Tiktok [53, 44], or the most recent trend of "What are you?" quiz/filters on Insta- gram [102]). Yet, I hope we find more ways to help people use communication technology to create more interactions that feel spontaneous and freeing.

6.3 Closing Statement

By understanding the ways that youth engage with and are influenced by social media technology, I was able to uncover some critical insights in the ways that technology can better support both online and in-person socializing with social media. There are opportunities for technology to further explore designing for a curated audience and a collaborative video creation experience. As a researcher, I am honored to bring insight and understanding to youths as prominent yet under-supported participants on the Internet. I believe the more that we can work together to build a safe and empowering place for our young users the better that the Internet will be for everyone. 85

Appendix A

Novel Prototype Development

In this appendix I introduce my novel prototype, Squid. Squid is a mobile app designed to support video creation among people in the same place, but with their own devices. Videos are created by one, short, segment at a time to en- courage collaboration and play in the video creation process.

A.1 Design Values

There are three motivating values in the design of this system: play, creativ- ity, and friendship. These values inform the design decisions throughout the application. Play. This design is more focused on the process of video creation than the final product. Previous work supports the idea that much of video sharing is about the fun of creating rather than getting a video perfect (esp. [70]). Creativity. This prototype is designed to help people split roles without needing them to be clearly defined. The collaborative system gives kids space Appendix A. Novel Prototype Development 86 to express their own idea on their own phones and then contribute it to a larger whole. Friendship. I intend to show that this prototype helps kids engage in an activity together with technology rather than necessarily through technology. The way that youth only engage with one another’s videos by adding to them, will contribute to a more collaborative environment for social media. I seek to support creation in a way that more people can contribute and bond with their peers.

A.1.1 Design Overview

We chose to develop a mobile application to recognize the popularity of mo- bile applications for sharing [62]. We leveraged primarily Google services to implement the application. We used Flutter (flutter.io) to develop the front end features. Flutter is a framework for native mobile development with built in features and packages to develop the interface, camera, video, and QR func- tionality. Button and interaction designs are modeled off of industry standards. We connected the backend with Firebase (firebase.google.com). This meant we were able to use a streamlined system for user log-in, video processing, and storage. The focus of backend work is storing and managing the individual videos made at each step as they relate to the larger collection of videos (re- ferred to as a "project") to which the creators are contributing, this data model is outlined in Figure A.1. Appendix A. Novel Prototype Development 87

FIGURE A.1: Squid is designed to help people focus on one video at a time in one project at a time. Each project is the specific iter- ation of videos that was created with the most recent video. We keep track of the previous version with the parentID. While there could be many videos in a project, one video (especially the first one) could be in many projects as a base that is built off of by many growing projects. Appendix A. Novel Prototype Development 88

A.2 Front End Features

This section details the experience of using Squid. Each step of project creation is designed to support the creative process.

A.2.1 Video Recording

This first step in using Squid is to create a video. Users can only add one clip to any video project before sharing it with the rest of the group to continue collaborations. This keeps the rest of the group involved and helps move along the video process. Other work shows the value of encouraging sharing works in progress (e.g. [57, 58]), and creating short videos follows a standard in popular, video apps. This feature helps kids focus more on creating and sharing between their friends. This has the added benefit of keeping the app lightweight for prototyping speed and back end memory loads. Chance to retake clip that will be inserted.

A.2.2 Video Passing

I designed the system of sharing of videos to explore the idea of creating physi- cally together. Other work has shown the potential of leveraging mobile phones for collaboration [66]. Squid generates a QR code that people have to scan with their app to access a friend’s video. After a video is recorded and the user is ready to share it with a friend, they will navigate to the next screen which will generate a QR code specific to that exact video. They’ll be able to share this QR code with nearby friends, who will scan the screen with the app on their phone. Appendix A. Novel Prototype Development 89

Scanning the code will allow them to start a new version of the project. Future work could investigate the potential of starting together, but completing a video separately (such as scanning a QR code to gain access to a video, but being able to return to it later through the app), or allowing people to come back to keep adding to a project.

A.2.3 Video Combining

Another distinct feature of Squid is the way that videos combine. Users have the chance to add their own video content to parts of the beginning, middle, or end of a video that was passed to them. This is also the only way to edit a video project. From our experience at Sanford, it can be distracting to have too many editing features to explore and manage. Limiting video modification to this one feature will contribute to the playfulness and collaborative spirit of Squid. Also, as my work has shown, the ability to remix is also a popular and under-supported feature in video creation [108, 69].

A.3 Back End Video Processing and Storage

The main technical challenge of this work is the snowball process of combin- ing videos. Not only are videos concatenating on a mobile device, but they’re also travelling from one device to another as the project grows. As Figure A.1 shows, each phone only contains information relevant to the current video they’re adding to the current project. Videos are stitched together at the end of Appendix A. Novel Prototype Development 90 their creation (either when a user archives the video or at the end of the ses- sion for study purposes). Until then, video segments are presented together in a playlist that mimics the look of a fully merged video. We manage the videos and projects separately to reduce redundancy. As shown in Figure A.2, each indi- vidual video is stored in Firebase Storage, while the meta data of each project is stored in Firestore (a flexible, scalable NoSQL cloud database). For each project, the Firestore will keep track of the order (and start and end times) of videos which are downloaded on the phone from Firebase Storage. This allows users to feel the experience of combining their videos without having to wait for videos to be edited, encoded, and redownloaded.

A.4 Conclusion

Squid is a mobile application that streamlines the video creation process to sup- port play, creativity, and friendship. These features are inspired by formative work in Chapters ??. The system leverages Flutter and Firebase to build an ex- perience that mimics the effect editing videos together into one collaborative piece. This prototype serves as the foundation for my proposed study. Appendix A. Novel Prototype Development 91

FIGURE A.2: The technical process for combining videos will em- ploy a “fake it ’til you make it” approach to combining videos. Rather than editing videos together, each video project will func- tion as a playlist that cuts between individually stored videos. The videos will be stored in the backend Firestore Storage, with the project metadata stored and processed in the Firestore database. 92

Bibliography

[1] URL: https://musical.ly/en-US/about.

[2] URL: https://boardgamegeek.com/boardgame/46213/telestrations.

[3] URL: https://www.snapchat.com/create.

[4] URL: http://attrakdiff.de/index-en.html.

[5] A vermont politician wants to ban kids from having cellphones. URL: https://www.mic.com/p/vermont- politician- proposes-

cellphone-ban-for-those-under-21-19783813.

[6] D. Abrahamson and R. Lindgren. Embodiment and embodied design. In Cambridge handbook of the learning sciences), part 18, pages 358–376. Cam- bridge University Press, 2nd edition, 2014.

[7] S. Alhabash and M. Ma. A tale of four platforms: motivations and uses of facebook, twitter, instagram, and snapchat among college students?

Social Media + Society, 3(1):2056305117691544, 2017. ISSN: 2056-3051. DOI: 10.1177/2056305117691544.

[8] T. Ammari, P. Kumar, C. Lampe, and S. Schoenebeck. Managing chil- dren’s online identities: how parents decide what to disclose about their BIBLIOGRAPHY 93

children online. In Proceedings of the 33rd Annual ACM Conference on Hu-

man Factors in Computing Systems, 1895–1904. ACM, 2015. ISBN: 978-1-

4503-3145-6. DOI: 10.1145/2702123.2702325. URL: http://doi. acm.org/10.1145/2702123.2702325.

[9] L. Anthony, Y. Kim, and L. Findlater. Analyzing user-generated videos to understand touchscreen use by people with motor impairments. In Proceedings of the SIGCHI Conference on Human Factors in Computing

Systems, 1223–1232. ACM, 2013. ISBN: 978-1-4503-1899-0. DOI: 10.1145/

2470654.2466158. URL: http://doi.acm.org/10.1145/2470654. 2466158.

[10] A. Antle. Case study: the design of cbc4kids’ storybuilder. In Proceedings of the 2003 Conference on Interaction Design and Children, IDC ’03, 59–68.

ACM, 2003. ISBN: 1-58113-732-X. DOI: 10.1145/953536.953546. URL: http://doi.acm.org/10.1145/953536.953546.

[11] C. R. Aragon, S. S. Poon, A. Monroy-Hernández, and D. Aragon. A tale of two online communities: fostering collaboration and creativity in sci- entists and children. In Proceedings of the Seventh ACM Conference on Cre-

ativity and Cognition, C&C ’09, 9–18. ACM, 2009. ISBN: 978-1-60558-865-0.

DOI: 10.1145/1640233.1640239. URL: http://doi.acm.org/ 10.1145/1640233.1640239.

[12] M. Asselin, T. Dobson, E. M. Meyers, C. Teixiera, and L. Ham. Learning from youtube: an analysis of information literacy in user discourse. In BIBLIOGRAPHY 94

Proceedings of the 2011 iConference, 640–642. ACM, 2011. ISBN: 978-1-4503-

0121-3. DOI: 10.1145/1940761.1940851. URL: http://doi.acm. org/10.1145/1940761.1940851.

[13] J. B. Bayer, N. B. Ellison, S. Y. Schoenebeck, and E. B. Falk. Sharing the small moments: ephemeral social interaction on snapchat. Information,

Communication & Society, 19(7):956–977, 2016. ISSN: 1369-118X. DOI: 10. 1080/1369118X.2015.1084349.

[14] C. Bell, C. Fausset, S. Farmer, J. Nguyen, L. Harley, and W. B. Fain. Ex- amining social media use among older adults. In Proceedings of the 24th ACM Conference on Hypertext and Social Media, HT ’13, 158–163. ACM,

2013. ISBN: 978-1-4503-1967-6. DOI: 10.1145/2481492.2481509. URL: http://doi.acm.org/10.1145/2481492.2481509.

[15] L. Benton, A. Vasalou, D. Gooch, and R. Khaled. Understanding and fostering children’s storytelling during game narrative design. In Pro- ceedings of the 2014 Conference on Interaction Design and Children, IDC ’14,

301–304. ACM, 2014. ISBN: 978-1-4503-2272-0. DOI: 10.1145/2593968.

2610477. URL: http://doi.acm.org/10.1145/2593968.2610477.

[16] E. Bonsignore, A. J. Quinn, A. Druin, and B. B. Bederson. Sharing stories “in the wild”: a mobile storytelling case study using storykit. ACM Trans.

Comput.-Hum. Interact., 20(3):18:1–18:38, 2013. ISSN: 1073-0516. DOI: 10. 1145/2491500.2491506.

[17] E. Brockes. Parents are exploiting their children on youtube for fame and

easy money | emma brockes. The Guardian, 2019. ISSN: 0261-3077. URL: BIBLIOGRAPHY 95

https://www.theguardian.com/commentisfree/2019/mar/ 22 / parents - exploiting - children - youtube - fame - easy -

money.

[18] T. Buhler, C. Neustaedter, and S. Hillman. How and why teenagers use video chat. In Proceedings of the 2013 Conference on Computer Supported

Cooperative Work, 759–768. ACM, 2013. ISBN: 978-1-4503-1331-5. DOI: 10.

1145/2441776.2441861. URL: http://doi.acm.org/10.1145/ 2441776.2441861.

[19] T. Buhler, C. Neustaedter, and S. Hillman. How and why teenagers use video chat. In Proceedings of the 2013 Conference on Computer Supported

Cooperative Work, 759–768. ACM, 2013. ISBN: 978-1-4503-1331-5. DOI: 10.

1145/2441776.2441861. URL: http://doi.acm.org/10.1145/ 2441776.2441861.

[20] T. Buhler, C. Neustaedter, and S. Hillman. How and why teenagers use video chat. In Proceedings of the 2013 Conference on Computer Supported

Cooperative Work, CSCW ’13, 759–768. ACM, 2013. ISBN: 978-1-4503-1331-

5. DOI: 10.1145/2441776.2441861. URL: http://doi.acm.org/ 10.1145/2441776.2441861.

[21] J. Burgess, J. Green, H. Jenkins, and J. Hartley. YouTube: Online Video

and Participatory Culture. Polity, 1 edition edition, 2009. ISBN: 978-0-7456- 4479-0. BIBLIOGRAPHY 96

[22] M. Buzzi. Children and youtube: access to safe content. In Proceedings of the 9th ACM SIGCHI Italian Chapter International Conference on Computer-

Human Interaction: Facing Complexity, CHItaly, 125–131. ACM, 2011. ISBN:

978-1-4503-0876-2. DOI: 10 . 1145 / 2037296 . 2037328. URL: http : //doi.acm.org/10.1145/2037296.2037328.

[23] R. C.-J. R. Cellan-Jones D. L. Under-age social media use “on the rise”.

BBC News, 2017. URL: http://www.bbc.com/news/technology- 42153694.

[24] P. Cesar, D. C. Bulterman, D. Geerts, J. Jansen, H. Knoche, and W. Sea- ger. Enhancing social sharing of videos: fragment, annotate, enrich, and share. In Proceedings of the 16th ACM International Conference on Multi-

media, MM ’08, 11–20. ACM, 2008. ISBN: 978-1-60558-303-7. DOI: 10 .

1145/1459359.1459362. URL: http://doi.acm.org/10.1145/ 1459359.1459362.

[25] D. Chandler and B. Torbert. Transforming inquiry and action: interweav-

ing 27 flavors of action research. Action Research, 1(2):133–152, 2003. ISSN:

1476-7503. DOI: 10.1177/14767503030012002.

[26] C. Chau. Youtube as a participatory culture. New Directions for Youth De-

velopment, 2010(128):65–74, 2010. ISSN: 1537-5781. DOI: 10.1002/yd. 376.

[27] C. Chau. YouTube as a participatory culture. en. New Directions for Youth

Development, 2010(128):65–74, Dec. 2010. ISSN: 1537-5781. DOI: 10.1002/ BIBLIOGRAPHY 97

yd . 376. URL: http : / / onlinelibrary . wiley . com / doi / 10 . 1002/yd.376/abstract (visited on 04/19/2016).

[28] Childnet. Age restrictions on social media services. URL: http://www. childnet.com/blog/age-restrictions-on-social-media-

services.

[29] S. L. Chu and F. Quek. Things to imagine with: designing for the child’s creativity. In Proceedings of the 12th International Conference on Interaction

Design and Children, IDC ’13, 261–264. ACM, 2013. ISBN: 978-1-4503-1918-

8. DOI: 10.1145/2485760.2485793. URL: http://doi.acm.org/ 10.1145/2485760.2485793.

[30] J. Constine. Facebook messenger day launches as a snapchat stories clone

for making plans. URL: http://social.techcrunch.com/2017/ 03/09/facebook-messenger-day/.

[31] K. Diehl, G. Zauberman, and A. Barasch. How taking photos increases enjoyment of experiences. Journal of Personality and Social Psychology, 111(2):119–140,

2016. ISSN: 1939-1315. DOI: 10.1037/pspa0000055.

[32] A. Druin. Cooperative inquiry: developing new technologies for children with children. In Proceedings of the SIGCHI Conference on Human Factors in

Computing Systems, CHI ’99, 592–599. ACM, 1999. ISBN: 978-0-201-48559-

2. DOI: 10.1145/302979.303166. URL: http://doi.acm.org/10. 1145/302979.303166.

[33] S. Egelman, E. H. Chi, and S. Dow. Crowdsourcing in hci research. In Ways of Knowing in HCI. J. S. Olson and W. A. Kellogg, editors. Springer New BIBLIOGRAPHY 98

York, 2014, 267–289. ISBN: 978-1-4939-0377-1, 978-1-4939-0378-8. URL: http: //link.springer.com/chapter/10.1007/978-1-4939-0378-

8_11.

[34] Facebook and instagram change to crack down on underage children.

URL: http://social.techcrunch.com/2018/07/19/facebok- under-13/.

[35] D. Fitton and J. C. Read. Creating a framework to support the critical consideration of dark design aspects in free-to-play apps. In Proceedings of the 18th ACM International Conference on Interaction Design and Children,

IDC ’19, 407–418. Association for Computing Machinery, 2019. ISBN: 978-

1-4503-6690-8. DOI: 10.1145/3311927.3323136. URL: http://doi. org/10.1145/3311927.3323136.

[36] M. Foth and J. Axup. Participatory design and action research: identi- cal twins or synergetic pair? In G. Jacucci, F. Kensing, I. Wagner, and J.

Blomberg, editors, Creative Industries Faculty, volume 2, 93–96, 2006. ISBN:

978-0-9667818-4-7. URL: http://eprints.qut.edu.au/4347/.

[37] S. Franckel, E. Bonsignore, and A. Druin. Designing for children’s mobile storytelling: International Journal of Mobile Human Computer Interaction,

2(2):19–36, 2010. ISSN: 1942-390X, 1942-3918. DOI: 10 . 4018 / jmhci . 2010040102.

[38] J. P.Gee, A.-R. Allen, and K. Clinton. Language, class, and identity: teenagers fashioning themselves through language. Linguistics and Education, 12(2):175–194,

2001. ISSN: 0898-5898. DOI: 10.1016/S0898-5898(00)00045-0. BIBLIOGRAPHY 99

[39] C. Grevet and E. Gilbert. Piggyback prototyping: using existing, large- scale social computing systems to prototype new ones. In Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems,

CHI ’15, 4047–4056. Association for Computing Machinery, 2015. ISBN:

978-1-4503-3145-6. DOI: 10 . 1145 / 2702123 . 2702395. URL: http : //doi.org/10.1145/2702123.2702395.

[40] R. E. Grinter and L. Palen. Instant messaging in teen life. In Proceedings of the 2002 ACM Conference on Computer Supported Cooperative Work, 21–30.

ACM, 2002. ISBN: 1-58113-560-2. DOI: 10.1145/587078.587082. URL: http://doi.acm.org/10.1145/587078.587082.

[41] M. L. Guha, A. Druin, and J. A. Fails. Cooperative inquiry revisited: re- flections of the past and guidelines for the future of intergenerational co-design. International Journal of Child-Computer Interaction, 1(1):14–23,

2013. ISSN: 2212-8689. DOI: 10.1016/j.ijcci.2012.08.003.

[42] E. I. K. Hansen and O. S. Iversen. You are the real experts!: studying teenagers’ motivation in participatory design. In Proceedings of the 12th International Conference on Interaction Design and Children, IDC ’13, 328–331.

ACM, 2013. ISBN: 978-1-4503-1918-8. DOI: 10.1145/2485760.2485826.

URL: http://doi.acm.org/10.1145/2485760.2485826.

[43] M. Hassenzahl. Hedonic, emotional, and experiential perspectives on product quality. Encyclopedia of Human Computer Interaction:266–272, 2006.

DOI: 10.4018/978-1-59140-562-7.ch042. BIBLIOGRAPHY 100

[44] J. Herrman. How tiktok is rewriting the world. The New York Times, 2019.

ISSN: 0362-4331. URL: https://www.nytimes.com/2019/03/10/ style/what-is-tik-tok.html.

[45] B. Hogan, A. Quan-Haase, B. Hogan, and A. Quan-Haase. Persistence and change in social media. Bulletin of Science, Technology Society, 30(5):309–315,

2010. ISSN: 0270-4676.

[46] J. P. Hourcade, S. L. Mascher, D. Wu, and L. Pantoja. Look, my baby is using an ipad! an analysis of youtube videos of infants and toddlers using tablets. In Proceedings of the 33rd Annual ACM Conference on Human

Factors in Computing Systems, 1915–1924. ACM, 2015. ISBN: 978-1-4503-

3145-6. DOI: 10.1145/2702123.2702266. URL: http://doi.acm. org/10.1145/2702123.2702266.

[47] G. A. Hull. At last: youth culture and digital media: new literacies for

new times. Research in the Teaching of English, 38(2):229–233, 2003. ISSN: 0034-527X.

[48] K. Inkpen, H. Du, A. Hoff, P. Johns, and A. Roseway. Video kids: aug- menting close friendships with asynchronous video conversations in videopal. In Proc. of CHI, 2388–2396, 2012.

[49] Inside facebook. URL: http://www.channel4.com/programmes/ inside-facebook-secrets-of-a-social-network.

[50] Introducing the newest member of our family, the youtube kids app–

available on google play and the app store, 2015. URL: http://youtube- global.blogspot.com/2015/02/youtube-kids.html. BIBLIOGRAPHY 101

[51] M. Isaac. Facebook mirrors snapchat (again) with stories feature. The

New York Times, 2017. ISSN: 0362-4331. URL: https://www.nytimes. com/2017/03/28/technology/facebook-mirrors-snapchat-

again-with-facebook-stories.html.

[52] M. Isaac. Instagram takes a page from snapchat, and takes aim at it,

too. The New York Times, 2016. ISSN: 0362-4331. URL: https : / / www . nytimes.com/2016/08/03/technology/instagram-stories-

snapchat-facebook.html.

[53] It’s time to pay serious attention to tiktok. TechCrunch. URL: http:// social.techcrunch.com/2019/01/29/its- time- to- pay-

serious-attention-to-tiktok/.

[54] O. S. Iversen, R. C. Smith, and C. Dindler. Child as protagonist: expand- ing the role of children in participatory design. In Proceedings of the 2017 Conference on Interaction Design and Children, IDC ’17, 27–37. ACM, 2017.

ISBN: 978-1-4503-4921-5. DOI: 10.1145/3078072.3079725. URL: http: //doi.acm.org/10.1145/3078072.3079725.

[55] H. Jia, P. J. Wisniewski, H. Xu, M. B. Rosson, and J. M. Carroll. Risk- taking As a Learning Process for Shaping Teen’s Online Information Pri- vacy Behaviors. In Proceedings of the 18th ACM Conference on Computer Supported Cooperative Work & Social Computing, CSCW ’15, pages 583–

599, New York, NY, USA. ACM, 2015. ISBN: 978-1-4503-2922-4. DOI: 10.

1145/2675133.2675287. URL: http://doi.acm.org/10.1145/ 2675133.2675287 (visited on 04/06/2016). BIBLIOGRAPHY 102

[56] C. Jones. Should children watch toy unboxing videos? BBC News, 2019.

URL: https://www.bbc.com/news/uk-england-beds-bucks- herts-49975644.

[57] J. Kim, M. Agrawala, and M. S. Bernstein. Mosaic: designing online cre- ative communities for sharing works-in-progress. In Proceedings of the 2017 ACM Conference on Computer Supported Cooperative Work and Social

Computing, CSCW ’17, 246–258. ACM, 2017. ISBN: 978-1-4503-4335-0. DOI:

10.1145/2998181.2998195. URL: http://doi.acm.org/10. 1145/2998181.2998195. event-place: Portland, Oregon, USA.

[58] J. Kim, S. Sterman, A. A. B. Cohen, and M. S. Bernstein. Mechanical novel: crowdsourcing complex work through reflection and revision. In Proceedings of the 2017 ACM Conference on Computer Supported Coopera-

tive Work and Social Computing, CSCW ’17, 233–245. ACM, 2017. ISBN:

978-1-4503-4335-0. DOI: 10 . 1145 / 2998181 . 2998196. URL: http : //doi.acm.org/10.1145/2998181.2998196. event-place: Port- land, Oregon, USA.

[59] L. J. Knoll, L. Magis-Weinberg, M. Speekenbrink, and S.-J. Blakemore. Social influence on risk perception during adolescence. Psychological Sci-

ence, 26(5):583–592, 2015. ISSN: 0956-7976. DOI: 10.1177/0956797615569578.

[60] R. Laiola Guimarães, P.Cesar, D. C. Bulterman, V.Zsombori, and I. Kegel. Creating personalized memories from social events: community-based support for multi-camera recordings of school concerts. In Proceedings of BIBLIOGRAPHY 103

the 19th ACM international conference on Multimedia, MM ’11, 303–312. As-

sociation for Computing Machinery, 2011. ISBN: 978-1-4503-0616-4. DOI:

10.1145/2072298.2072339. URL: http://doi.org/10.1145/ 2072298.2072339.

[61] A. Leavitt. "This is a Throwaway Account": Temporary Technical Iden- tities and Perceptions of Anonymity in a Massive Online Community. In Proceedings of the 18th ACM Conference on Computer Supported Coopera- tive Work & Social Computing, CSCW ’15, pages 317–327, New York, NY,

USA. ACM, 2015. ISBN: 978-1-4503-2922-4. DOI: 10.1145/2675133.

2675175. URL: http://doi.acm.org/10.1145/2675133.2675175 (visited on 05/02/2016).

[62] A. Lenhart. Teens, social media & technology overview 2015, 2015. URL: http://www.pewinternet.org/2015/04/09/teens-social-

media-technology-2015/.

[63] T. Li, K. Luther, and C. North. Crowdia: solving mysteries with crowd- sourced sensemaking. Proceedings of the ACM on Human-Computer Inter-

action, 2(CSCW):105:1–105:29, 2018. DOI: 10.1145/3274374.

[64] E. Litt. Knock, knock. who’s there? the imagined audience. Journal of

Broadcasting & Electronic Media, 56(3):330–345, 2012. ISSN: 0883-8151. DOI: 10.1080/08838151.2012.705195. BIBLIOGRAPHY 104

[65] S. Livingstone. Taking risky opportunities in youthful content creation: teenagers’ use of social networking sites for intimacy, privacy and self-

expression. New Media & Society, 10(3):393–411, 2008. ISSN: 1461-4448,

1461-7315. DOI: 10.1177/1461444808089415.

[66] A. Lucero, J. Keränen, and H. Korhonen. Collaborative use of mobile phones for brainstorming. In Proceedings of the 12th International Con- ference on Human Computer Interaction with Mobile Devices and Services,

MobileHCI ’10, 337–340. ACM, 2010. ISBN: 978-1-60558-835-3. DOI: 10.

1145/1851600.1851659. URL: http://doi.acm.org/10.1145/ 1851600.1851659. event-place: Lisbon, .

[67] K. Luther, C. Fiesler, and A. Bruckman. Redistributing leadership in on- line creative collaboration. In Proceedings of the 2013 Conference on Com-

puter Supported Cooperative Work, CSCW ’13, 1007–1022. ACM, 2013. ISBN:

978-1-4503-1331-5. DOI: 10 . 1145 / 2441776 . 2441891. URL: http : //doi.acm.org/10.1145/2441776.2441891.

[68] J. McNiff. Action Research: Principles and Practice. Routledge, 2013. ISBN: 978-0-415-53525-0. Google-Books-ID: wCTzD9PMQ_0C.

[69] S. McRoberts, E. Bonsignore, T. Peyton, and S. Yarosh. Do it for the view- ers!: audience engagement behaviors of young youtubers. In Proceed- ings of the The 15th International Conference on Interaction Design and Chil-

dren, IDC ’16, 334–343. ACM, 2016. ISBN: 978-1-4503-4313-8. DOI: 10 .

1145/2930674.2930676. URL: http://doi.acm.org/10.1145/ 2930674.2930676. BIBLIOGRAPHY 105

[70] S. McRoberts, H. Ma, A. Hall, and S. Yarosh. Share first, save later: per- formance of self through snapchat stories. In To Appear in Proceedings of CHI 2017, 2017.

[71] M. Muller. Curiosity, creativity, and surprise as analytic tools: grounded the- ory method. In Ways of Knowing in HCI. J. S. Olson and W. A. Kellogg,

editors. Springer New York, 2014, 25–48. ISBN: 978-1-4939-0377-1. DOI:

10.1007/978-1-4939-0378-8_2. URL: http://link.springer. com/chapter/10.1007/978-1-4939-0378-8_2.

[72] B. Nansen. Accidental, assisted, automated: an emerging repertoire of

infant mobile media techniques. M/C Journal, 18(5), 2015. ISSN: 14412616.

URL: http://www.journal.media- culture.org.au/index. php/mcjournal/article/view/1026.

[73] A.-C. News. Lawmaker wants to bar kids under 13 from using social

media, set half-hour limit. URL: https://news.abs-cbn.com/news/ 11/14/19/lawmaker-wants-to-bar-kids-under-13-from-

using-social-media-set-half-hour-limit.

[74] K. M.c.M.c. b. news and featuresEmailEmailBioBioFollowFollow. A law- maker wants to end ‘social media addiction’ by killing features that en-

able mindless scrolling. URL: https://www.washingtonpost.com/ technology / 2019 / 07 / 30 / lawmaker - wants - end - social - media - addiction - by - killing - features - that - enable -

mindless-scrolling/. BIBLIOGRAPHY 106

[75] H. Niemi and J. Multisilta. Digital storytelling promoting twenty-first century skills and student engagement. Technology, Pedagogy and Educa-

tion, 25(4):451–468, 2016. ISSN: 1475-939X. DOI: 10.1080/1475939X. 2015.1074610.

[76] L. Piwek and A. Joinson. “what do they snapchat about?” patterns of use in time-limited instant messaging service. Computers in Human Behavior,

54:358–367, 2016. ISSN: 0747-5632. DOI: 10.1016/j.chb.2015.08. 026.

[77] K. Purcell. Online video 2013. URL: http://www.pewinternet.org/ 2013/10/10/online-video-2013/.

[78] L. Rainie, J. Brenner, and K. Purcell. Photos and videos as social cur-

rency online. URL: http://www.pewinternet.org/2012/09/13/ photos-and-videos-as-social-currency-online/.

[79] Revisiting the privacy paradox. URL: https://dl.acm.org/citation. cfm?id=3217907.

[80] S. Rintel, R. Harper, and K. O’Hara. The Tyranny of the Everyday in Mo- bile Video Messaging. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems, CHI ’16, pages 4781–4792, New York, NY,

USA. ACM, 2016. ISBN: 978-1-4503-3362-7. DOI: 10.1145/2858036.

2858042. URL: http://doi.acm.org/10.1145/2858036.2858042 (visited on 09/19/2016). BIBLIOGRAPHY 107

[81] S. Rintel, R. Harper, and K. O’Hara. The tyranny of the everyday in mo- bile video messaging. In Proceedings of the 2016 CHI Conference on Hu-

man Factors in Computing Systems, CHI ’16, 4781–4792. ACM, 2016. ISBN:

978-1-4503-3362-7. DOI: 10 . 1145 / 2858036 . 2858042. URL: http : //doi.acm.org/10.1145/2858036.2858042.

[82] F. Roesner, B. T. Gill, and T. Kohno. Sex, Lies, or Kittens? Investigating the Use of Snapchat’s Self-Destructing Messages. en. In N. Christin and R. Safavi-Naini, editors, Financial Cryptography and Data Security, num- ber 8437 in Lecture Notes in Computer Science, pages 64–76. Springer

Berlin Heidelberg, Mar. 2014. ISBN: 978-3-662-45471-8 978-3-662-45472-5.

URL: http://link.springer.com/chapter/10.1007/978-3- 662-45472-5_5 (visited on 06/17/2015).

[83] C. Sas, S. Whittaker, S. Dow, J. Forlizzi, and J. Zimmerman. Generat- ing implications for design through design research. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, CHI ’14,

1971–1980. ACM, 2014. ISBN: 978-1-4503-2473-1. DOI: 10.1145/2556288.

2557357. URL: http://doi.acm.org/10.1145/2556288.2557357.

[84] J. Sauer and A. Sonderegger. The influence of prototype fidelity and aes- thetics of design in usability tests: effects on user behaviour, subjective

evaluation and emotion. Applied Ergonomics, 40(4):670–677, 2009. ISSN:

00036870. DOI: 10.1016/j.apergo.2008.06.006. BIBLIOGRAPHY 108

[85] B. Settles and S. Dow. Let’s get together: the formation and success of on- line creative collaborations. In Proceedings of the SIGCHI Conference on Hu-

man Factors in Computing Systems, CHI ’13, 2009–2018. ACM, 2013. ISBN:

978-1-4503-1899-0. DOI: 10 . 1145 / 2470654 . 2466266. URL: http : //doi.acm.org/10.1145/2470654.2466266.

[86] M. Sharmin and B. P. Bailey. “i reflect to improve my design”: investi- gating the role and process of reflection in creative design. In Proceedings of the 8th ACM Conference on Creativity and Cognition, C&C ’11, 389–390.

ACM, 2011. ISBN: 978-1-4503-0820-5. DOI: 10.1145/2069618.2069710.

URL: http://doi.acm.org/10.1145/2069618.2069710.

[87] A. Smith, Monica, and erson. Social media use in 2018, 2018. URL: http: //www.pewinternet.org/2018/03/01/social- media- use-

in-2018/.

[88] Snap inc. URL: https://www.snap.com/en-US/.

[89] Social media usage: 2005-2015. URL: http : / / www . pewinternet . org/2015/10/08/social-networking-usage-2005-2015/.

[90] C. Sylla. Designing a tangible interface for collaborative storytelling to access “embodiment” and meaning making. In Proceedings of the 12th In- ternational Conference on Interaction Design and Children, IDC ’13, 651–654.

ACM, 2013. ISBN: 978-1-4503-1918-8. DOI: 10.1145/2485760.2485881.

URL: http://doi.acm.org/10.1145/2485760.2485881. BIBLIOGRAPHY 109

[91] L. Taber and S. Whittaker. Personality depends on the medium: differ- ences in self-perception on snapchat, facebook and offline. In Proceed- ings of the 2018 CHI Conference on Human Factors in Computing Systems,

CHI ’18, 607:1–607:13. ACM, 2018. ISBN: 978-1-4503-5620-6. DOI: 10 .

1145/3173574.3174181. URL: http://doi.acm.org/10.1145/ 3173574.3174181.

[92] J. C. Tang, G. Venolia, and K. M. Inkpen. Meerkat and periscope: i stream, you stream, apps stream for live streams. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems, CHI ’16, 4770–4780.

ACM, 2016. ISBN: 978-1-4503-3362-7. DOI: 10.1145/2858036.2858374.

URL: http://doi.acm.org/10.1145/2858036.2858374.

[93] The bogus “momo challenge” internet , explained. URL: https : //www.vox.com/2019/3/3/18248783/momo-challenge-hoax-

explained.

[94] L. Van Leijenhorst, B. G. Moor, Z. A. O. d. Macks, S. A.R. B. Rombouts, P. M. Westenberg, and E. A. Crone. Adolescent risky decision-making: Neurocognitive development of reward and control regions. NeuroImage,

51(1):345–355, May 2010. ISSN: 1053-8119. DOI: 10.1016/j.neuroimage.

2010.02.038. URL: http://www.sciencedirect.com/science/ article/pii/S1053811910002065 (visited on 09/17/2018).

[95] J. M. Vaterlaus, K. Barnett, C. Roche, and J. A. Young. “Snapchat is more personal”: An exploratory study on Snapchat behaviors and young adult interpersonal relationships. Computers in Human Behavior, 62:594–601, Sept. BIBLIOGRAPHY 110

2016. ISSN: 0747-5632. DOI: 10 . 1016 / j . chb . 2016 . 04 . 029. URL: http : / / www . sciencedirect . com / science / article / pii /

S0747563216303041 (visited on 08/30/2016).

[96] G. Venolia, J. C. Tang, and K. Inkpen. Seesaw: i see you saw my video message. In Proceedings of the 17th International Conference on Human-Computer Interaction with Mobile Devices and Services, MobileHCI ’15, 244–253. ACM,

2015. ISBN: 978-1-4503-3652-9. DOI: 10.1145/2785830.2785847. URL: http://doi.acm.org/10.1145/2785830.2785847.

[97] Vine is dead. URL: https://gizmodo.com/vine-is-dead-1788285242.

[98] G. Walsh, E. Foss, J. Yip, and A. Druin. Facit pd: a framework for analy- sis and creation of intergenerational techniques for participatory design. In Proceedings of the SIGCHI Conference on Human Factors in Computing

Systems, CHI ’13, 2893–2902. ACM, 2013. ISBN: 978-1-4503-1899-0. DOI:

10.1145/2470654.2481400. URL: http://doi.acm.org/10. 1145/2470654.2481400.

[99] G. Wang, B. Wang, T. Wang, A. Nika, H. Zheng, and B. Y. Zhao. Whis- pers in the Dark: Analysis of an Anonymous Social Network. In Proceed- ings of the 2014 Conference on Internet Measurement Conference, IMC ’14,

pages 137–150, New York, NY, USA. ACM, 2014. ISBN: 978-1-4503-3213-

2. DOI: 10.1145/2663716.2663728. URL: http://doi.acm.org/ 10.1145/2663716.2663728 (visited on 09/09/2015). BIBLIOGRAPHY 111

[100] A. Weibert and K. Schubert. How the social structure of intercultural computer clubs fosters interactive storytelling. In Proceedings of the 9th In- ternational Conference on Interaction Design and Children, IDC ’10, 368–371.

ACM, 2010. ISBN: 978-1-60558-951-0. DOI: 10.1145/1810543.1810616.

URL: http://doi.acm.org/10.1145/1810543.1810616.

[101] J. Wellemeyer. Youtube considers removing child stars — their parents

say that’s a drastic solution to a complex problem. URL: https : / / www.marketwatch.com/story/parents-of-youtube-child- stars - say - proposed - safety - protections - could - cost -

them-money-2019-06-26.

[102] Why are we obsessed with instagram’s ‘what are you?’ filter? Engad-

get. URL: https : / / www . engadget . com / 2020 / 02 / 03 / which - pokemon-are-you-instagram-filters-explained/.

[103] P. Wisniewski, H. Jia, N. Wang, S. Zheng, H. Xu, M. B. Rosson, and J. M. Carroll. Resilience mitigates the negative effects of adolescent internet addiction and online risk exposure. In Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems, CHI ’15, 4029–4038.

ACM, 2015. ISBN: 978-1-4503-3145-6. DOI: 10.1145/2702123.2702240.

URL: http://doi.acm.org/10.1145/2702123.2702240.

[104] P. Wisniewski, H. Xu, M. B. Rosson, D. F. Perkins, and J. M. Carroll. Dear Diary: Teens Reflect on Their Weekly Online Risk Experiences. In Proceed- ings of the 2016 CHI Conference on Human Factors in Computing Systems, BIBLIOGRAPHY 112

CHI ’16, pages 3919–3930, New York, NY, USA. ACM, 2016. ISBN: 978-1-

4503-3362-7. DOI: 10.1145/2858036.2858317. URL: http://doi. acm.org/10.1145/2858036.2858317 (visited on 05/12/2016).

[105] P. J. Wisniewski, H. Xu, M. B. Rosson, and J. M. Carroll. Adolescent on- line safety: the “moral” of the story. In Proc. of CSCW, 1258–1271. ACM, 2014.

[106] M. F. Wright. Perceptions of popularity-related behaviors in the cyber context: relations to cyber social behaviors. Child Development Research,

2015:e636929, 2015. ISSN: 2090-3987. DOI: 10.1155/2015/636929.

[107] B. Xu, P. Chang, C. L. Welker, N. N. Bazarova, and D. Cosley. Automatic archiving versus default deletion: what snapchat tells us about ephemer- ality in design. In Proceedings of the 19th ACM Conference on Computer- Supported Cooperative Work & Social Computing, CSCW ’16, 1662–1675.

ACM, 2016. ISBN: 978-1-4503-3592-8. DOI: 10.1145/2818048.2819948.

URL: http://doi.acm.org/10.1145/2818048.2819948.

[108] S. Yarosh, E. Bonsignore, S. McRoberts, and T. Peyton. Youthtube: youth video authorship on youtube and vine. In Proceedings of the 2016 ACM Conference on Computer Supported Cooperative Work. ACM, 2016.

[109] Y.-C. G. Yen, S. P. Dow, E. Gerber, and B. P. Bailey. Listen to others, listen to yourself: combining feedback review and reflection to improve itera- tive design. In Proceedings of the 2017 ACM SIGCHI Conference on Creativ-

ity and Cognition, C&C ’17, 158–170. ACM, 2017. ISBN: 978-1-4503-4403-6. BIBLIOGRAPHY 113

DOI: 10.1145/3059454.3059468. URL: http://doi.acm.org/ 10.1145/3059454.3059468.

[110] J. C. Yip, E. Foss, E. Bonsignore, M. L. Guha, L. Norooz, E. Rhodes, B. Mc- Nally, P. Papadatos, E. Golub, and A. Druin. Children initiating and lead- ing cooperative inquiry sessions. In Proceedings of the 12th International Conference on Interaction Design and Children, IDC ’13, 293–296. ACM,

2013. ISBN: 978-1-4503-1918-8. DOI: 10.1145/2485760.2485796. URL: http://doi.acm.org/10.1145/2485760.2485796.

[111] J. C. Yip, K. Sobel, C. Pitt, K. J. Lee, S. Chen, K. Nasu, and L. R. Pina. Examining adult-child interactions in intergenerational participatory de- sign. In Proceedings of the 2017 CHI Conference on Human Factors in Com-

puting Systems, CHI ’17, 5742–5754. ACM, 2017. ISBN: 978-1-4503-4655-9.

DOI: 10.1145/3025453.3025787. URL: http://doi.acm.org/ 10.1145/3025453.3025787.

[112] P.Yuksel-Arslan, S. Yildirim, and B. R. Robin. A phenomenological study: teachers’ experiences of using digital storytelling in early childhood edu-

cation. Educational Studies, 42(5):427–445, 2016. ISSN: 0305-5698. DOI: 10. 1080/03055698.2016.1195717.

[113] X. Zhao, N. Salehi, S. Naranjit, S. Alwaalan, S. Voida, and D. Cosley. The many faces of facebook: experiencing social media as performance, exhibition, and personal archive. In Proceedings of the SIGCHI Conference

on Human Factors in Computing Systems, CHI ’13, 1–10. ACM, 2013. ISBN: BIBLIOGRAPHY 114

978-1-4503-1899-0. DOI: 10 . 1145 / 2470654 . 2470656. URL: http : //doi.acm.org/10.1145/2470654.2470656.

[114] “two shots” of vodka. URL: https://knowyourmeme.com/videos/ 171932.