PaperlayerView exact revision on arXiv

Source-linked AI summary

SwitchTube: A Proof-of-Concept System Introducing "Adaptable Commitment Interfaces" as a Tool for Digital Wellbeing

Kai Lukoff, Ulrik Lyngs, Karina Shirokova, Raveena Rao, Larry Tian, Himanshu Zade, Sean A. Munson, Alexis Hiniker

arXiv:2610.22947v1cs.HC

TL;DR

Conventional digital self-control tools can block content users want, and it is unclear whether an easily switchable alternative can still support their goals. SwitchTube lets users toggle between search-first Focus and recommendations-first Explore modes; in a three-week field experiment, Switch generally supported agency, satisfaction, and goal alignment. The paper concludes that adaptable interfaces can support users’ efforts to shift the quality of their consumption.

  • Problem

    Conventional commitment interfaces can indiscriminately block wanted content, and it remains an open question whether easy switching undermines their commitment function.

    Source: §2.3 · p. 4

  • Method

    SwitchTube provided Focus, Explore, and toggleable Switch interfaces, which 46 U.S. participants used for one week each in a three-week field experiment.

    Source: §1 · p. 2

  • Results

    Switch better supported agency, satisfaction, and goal alignment than Explore alone, and offered greater satisfaction than Focus alone.

    Source: §6 · p. 18

  • Takeaways & Limitations

    SwitchTube demonstrates a digital wellbeing design that supports shifting the quality of video consumption as well as addressing time spent.

    Source: §7 · p. 20

    The study compared time spent in SwitchTube versions but did not measure whether use spilled over to YouTube itself.

    Source: §6.4 · p. 20

Abstract

from arXiv · show

YouTube has many features, such as homepage recommendations, that encourage users to explore its vast library of videos. However, when users visit YouTube with a specific intention, e.g., learning how to program in Python, these features to encourage exploration are often distracting. Prior work has innovated 'commitment interfaces' that restrict social media but finds that they often indiscriminately block needed content. In this paper, we describe the design, development, and evaluation of an 'adaptable commitment interface,' the SwitchTube mobile app, in which users can toggle between two interfaces when watching YouTube videos: Focus Mode (search-first) and Explore Mode (recommendations-first). In a three-week field deployment with 46 US participants, we evaluate how the ability to switch between interfaces affects user experience, finding that it provides users with a greater sense of agency, satisfaction, and goal alignment. We conclude with design implications for how adaptable commitment interfaces can support digital wellbeing.

1 INTRODUCTION

Media features can draw YouTube users away from their intentions, while conventional commitment tools may block content they still want. SwitchTube tests whether toggling between focused and exploratory interfaces can support users’ goals without unduly restricting exploration.

1 INTRODUCTION

  • Media use had a reported self-control failure rate of 76%, higher than smoking, drinking alcohol, or shopping in the cited study.The passage also reports failure rates of about 30% for smoking, 30% for drinking, and 35% for shopping.Source: §1 · p. 2
  • YouTube recommendations, autoplay, and related videos can draw users away from educational content or their intended stopping point, diminishing their sense of agency.Users may visit for educational videos but be shown more entertaining recommendations.Source: §1 · p. 2
  • Conventional commitment interfaces can block wanted content, while an adaptable interface raises the question of whether easy switching weakens commitment.The proposed alternative lets users toggle interfaces to fit situational needs.Source: §1 · p. 2
  • In a survey of 606 U.S. YouTube users, about half were trying or planning to change their use, and about half of their goals concerned video quality rather than quantity.About one-third of participant goals concerned reducing quantity of use.Source: §1 · p. 2
  • In a three-week field experiment, 46 participants used Explore, Focus, and SwitchTube’s toggleable Switch Version; Switch generally offered goal focus without severely restricting exploration.Users valued starting with focused search before exploring recommendations.Source: §1 · p. 2
  • The paper reports that adaptable commitment interfaces can improve short-term satisfaction and long-term goal alignment, and offers design implications for supporting agency.These are stated contributions of the paper.Source: §1 · pp. 2–3

2 BACKGROUND AND MOTIVATION

The background frames digital wellbeing around both what users watch and how much, then introduces adaptable commitment interfaces as a way to balance goal support with situational control. It concludes with questions about user goals, switching behavior, and experience.

2 BACKGROUND AND MOTIVATION

  • The background positions adaptable commitment interfaces as a design approach for addressing problematic YouTube use and the challenge of choosing enforcement strictness.It first reviews problematic YouTube use and then considers the opportunity and challenges of adaptable interfaces.Source: §2 · p. 3, §2.2 · p. 4

2.1 YouTube as a Site of Problematic Technology Use

  • YouTube can conflict with users’ preferences through both the quantity and quality of consumption, while existing screen-time tools mainly limit time rather than changing encountered content.Quality-related concerns include short-term content choice and long-term cultivation effects.Source: §2.1 · p. 3

2.2 Commitment Interfaces for Digital Self-Control

  • Commitment interfaces delegate a present self’s goal to a tool that constrains the future self, but enforcement must balance goal support against frustration and abandonment.Weak enforcement may be easy to circumvent, whereas overly strong enforcement can trigger reactance.Source: §2.2 · pp. 3–4

2.3 Adaptable Commitment Interfaces for Digital Wellbeing

  • An adaptable commitment interface lets users choose different levels of control to fit a situation, but easy access to a tempting alternative may weaken its commitment function.The section contrasts this risk with the possibility of giving users the interface that best suits their needs.Source: §2.3 · p. 4

2.4 Research Questions

  • The paper asks how U.S. YouTube users want to change consumption, when and why they switch interfaces, and how those interfaces affect satisfaction and goal alignment.The study reports the survey before the SwitchTube deployment addressing switching and experience.Source: §2.4 · pp. 4–5

3 SURVEY: STAGES OF CHANGE FOR YOUTUBE USERS

The survey used stages of change to examine whether U.S. YouTube users wanted to change their use and what changes they sought. Most reported goals spanned both time spent and content watched, motivating more granular design approaches than time limits alone.

3.1 Methods

  • The survey received 606 completed responses from YouTube users recruited in the U.S. who used Android phones and YouTube at least monthly.The sample included more young people than the general U.S. population.Source: §3.1 · p. 5

3.2 Procedure

  • A questionnaire adapted from behavior-change research classified participants’ readiness to change YouTube use, followed by an open-ended question about intended changes.Participants also rated how important the intended change was to them.Source: §3.2 · p. 5

3.3 Coding Reliability Thematic Analysis

  • Three authors developed a codebook from 100 responses, then two authors applied it to 50 more; agreement across codes and subcodes ranged from Cohen’s Kappa 0.66 to 1.The authors then divided the remaining responses and coded them using the shared codebook.Source: §3.3 · p. 5

3.4 Result and Analysis

  • 40% of participants were in Action and 45% in Maintenance, producing a bimodal distribution; about half were actively trying or preparing to change their YouTube use.The survey categorized participants into five stages, from Precontemplation to Action; the authors could not tell whether Maintenance reflected a prior change or satisfaction all along.Source: §3.4 · p. 6
  • Participants reported 817 goals: 37% sought less time on YouTube, while 28% sought more of some content and 23% sought less of some content.Among content goals, participants wanted more educational videos (34%) and less meaningless content (27%); recommendations accounted for 15% of goals to reduce.Source: §3.4 · p. 7
  • Pre-installed screen-time tools addressed about 40% of participants’ change goals, mainly by limiting time or presenting usage statistics, rather than supporting content changes within apps.The authors therefore aimed to address users’ needs through SwitchTube.Source: §3.4 · p. 7

4 SWITCHTUBE FIELD EXPERIMENT: EVALUATING AN ADAPTABLE COMMITMENT INTERFACE

The authors developed SwitchTube to test whether users could benefit from choosing between a search-first Focus interface and a recommendations-first Explore interface. They used co-design findings to motivate this adaptable approach, then specified hypotheses and evaluated three versions in a three-week field experiment.

4.1 Preparatory Design Work

  • Co-design sessions indicated that users wanted different levels of control in different situations, prompting Explore, Focus, and a Switch version that lets users toggle between them.Users preferred search-first for specific intentions and recommendations-first to relax or pass time; this led the authors to test whether Switch might provide the greatest agency.Source: §4.1 · p. 7
  • Explore added “viral” to search queries and showed recommendations, related videos, and autoplay; Focus hid recommendations and removed related videos and autoplay.The authors used non-personalized popular U.S. category videos because the YouTube Data API did not provide access to personalized recommendations.Source: §4.1 · p. 9
  • A YouTube Data API quota of 10,000 requests per day quickly ran out during testing, forcing the team to use a second API for search results.The authors report that this workaround required considerable time and effort.Source: §4.1 · p. 9

4.2 Pre-Registered Hypotheses

  • The authors preregistered hypotheses that Switch would provide the best combination of agency, satisfaction, and alignment with users’ long-term goals.They planned confirmatory experience-sampling analyses and descriptive-only exploratory analyses of app log data.Source: §4.2 · p. 9

4.3 Methods

  • Forty-six participants used Explore, Focus, and Switch for one week each; Switch always came last so participants could first experience both other interfaces.Participants had to use the app on at least three days and for 30 minutes per week; the study collected logs, experience-sampling responses, exit surveys, and interviews.Source: §4.3 · p. 11

4.4 Data Analysis

  • The analysis treated app logs descriptively, tested preregistered hypotheses with experience-sampling data, and used thematic coding of exit surveys and interviews.The qualitative analysis identified four switching use cases and organized findings around switching behavior, agency, satisfaction, and goal alignment.Source: §4.4 · p. 12

5 RESULTS AND ANALYSIS

The results and analysis examine how Focus, Explore, and switching between them shaped participants’ viewing experiences, drawing on app-use logs, ratings, and user reports. They cover intentional and discovery-oriented viewing, agency, satisfaction, goal alignment, and reported guilt about wasting time.

5.1 App Usage Data

  • Focus users made a median of 10 searches versus 5.5 in Explore and 4.5 in Switch, while Explore users watched 16.5 videos versus 11 in both other modes.Participants spent more time per started video in Focus, but spent more time browsing in Explore (49%) than Focus (46%) or Switch (38%); recommendations likely contributed to browsing in Explore.Source: §5.1 · p. 13
  • Explore users watched a median of 6 videos from homepage recommendations versus 1 in Focus, while search supplied 7.5 videos in Focus versus 2 in Explore.Related videos supplied a median of 3 watched videos in Explore and were unavailable in Focus; Switch fell between Explore and Focus across sources, indicating participants used both modes’ features.Source: §5.1 · p. 13, Figure 8 · p. 14
  • Explore selections were more diverse and often popular, while Focus selections tended to be niche videos related to the same search.Explore’s recommendation-driven viewing included broad-audience content; Focus’s search-driven viewing was more closely related to participants’ specific interests.Source: §5.1 · p. 13

5.2 The SwitchTube User Experience

  • Participants made a median of 8.5 switches during their week, using the toggle to compare features and discover preferences or to see novel recommendations.Recommendations sometimes inspired new searches, while switching between sessions let participants select Focus for educational or restrictive goals and Explore for easier entertainment or discovery.Source: §5.2.1 · p. 15
  • Participants also switched within sessions, starting in Focus to search without homepage distractions and then turning on Explore to see related videos.This unanticipated use case let participants begin with their intention and explore after a video started or finished.Source: §5.2.1 · p. 15, Figure 9 · p. 16
  • Experience-sampling ratings indicated greater sense of agency in Switch and Focus than Explore, consistent with the manipulation check: Focus features increased agency, while Explore features decreased it.Switch was rated higher than Focus, but not quite significantly so (p < 0.059); inability to customize Explore recommendations led participants to feel their attention had been led astray.Source: §5.2.2 · pp. 15–16, Figure 10 · p. 16
  • Satisfaction medians were 5.3 for Switch, 4.8 for Focus, and 4.2 for Explore, with significant differences across conditions and between every pair.Switch rated highest, while Focus unexpectedly rated higher than Explore; participants appreciated being able to switch to meet different needs, even when they did not use the option.Source: §5.2.3 · p. 17, Figure 11 · p. 17
  • Goal-alignment medians were 5.3 for Switch, 4.8 for Focus, and 4.3 for Explore; ratings were significantly higher for Switch and Focus than Explore.Focus’s minimal interface helped users avoid recommendations and related videos, focus on intended videos, and finish their goals; participants also described starting focused and then exploring within a session.Source: §5.2.4 · pp. 17–18, Figure 12 · p. 17
  • Using an initial video to direct exploration helped participants stick to their original goal and feel satisfied, even when they spent less time.Source: §5.2.4 · p. 18
  • Focus better supported participants than Explore even when they spent the same amount of time in each, because its content quality better aligned with their goals.Source: §5.2.4 · p. 18
  • Participants said Focus made it easier to feel satiated and less guilty about wasting time than Explore.Source: §5.2.4 · p. 18

6 DISCUSSION

The discussion finds that users seek changes in both YouTube time and content quality, and that Switch better supported agency, satisfaction, and goal alignment than Explore. It considers how adaptable interfaces can preserve choice while supporting users’ intentions, and identifies limits in the study’s recommendations and measures.

6 DISCUSSION

  • Switch better supported agency, satisfaction, and goal alignment than Explore, and offered greater satisfaction than Focus, despite more time spent in Focus than Explore.Median time was 145 minutes in Focus and 117 minutes in Explore; participants also switched between sessions to follow intentions and within sessions to search before exploring.Source: §6 · p. 18

6.1 Quality of Time, Not Just Quantity

  • Changing content quality accounted for 51% of survey goals, compared with 37% seeking to decrease time, and Switch participants rated Focus highly despite spending more time in it.Participants said time in Focus was usually better spent; the discussion cautions that screen time alone can overemphasize quantity as a measure of wellbeing.Source: §6.1 · p. 18

6.2 Adaptable Commitment Interfaces for Digital Wellbeing

  • An adaptable commitment interface can avoid distraction triggers while preserving users’ ability to switch, unlike restrictions that may block desired features or provoke reactance.The authors attribute Switch’s performance to avoiding triggers for distraction; they also recommend Focus Mode in multipurpose services such as YouTube, Twitter, and Facebook.Source: §6.2 · p. 19

6.3 From Adaptable to Adaptive Commitment Interfaces

  • A search-informed adaptive interface could use queries to infer intent, keeping specific-goal searches in Focus and shifting open-ended entertainment searches to Explore.This is proposed as a possible semi-automatic direction based on within-session switching; intent predictions can still be erroneous and may provoke reactance.Source: §6.3 · pp. 19–20

6.4 Limitations and Future Work

  • The study did not measure spillover to YouTube itself, and API limitations required workarounds to alter content quality in the study app.The authors also note that homepage recommendations were not personalized, limiting comparisons with YouTube and potentially affecting Explore ratings and use.Source: §6.1 · p. 18, §6.4 · p. 20

7 CONCLUSION

SwitchTube explores digital wellbeing beyond screen-time reduction by supporting changes in consumption quality. In a three-week field experiment with 46 US YouTube users, the Switch Version received the highest ratings, supporting adaptable commitment interfaces that let users focus or explore as they choose.

7 CONCLUSION

  • The Switch Version was rated highest of the three versions, while Focus rated higher than Explore on user-experience metrics without reducing time spent watching videos.The Android app was evaluated with 46 US YouTube users over three weeks.Source: §7 · p. 20
  • SwitchTube demonstrates how digital wellbeing design might support users in shifting consumption quality, not only reducing screen time.Source: §7 · p. 20
  • The authors recommend adaptable commitment interfaces that support users in exploring when they want to explore and focusing when they want to focus.Source: §7 · p. 20

Related papers

Loading 2610.22947v1…