key: cord-0265108-g64c0vzi
authors: Khan, Mina; Patel, Zeel; Wantlin, Kathryn; Glassman, Elena; Maes, Pattie
title: Changing Computer-Usage Behaviours: What Users Want, Use, and Experience
date: 2022-01-02
journal: nan
DOI: 10.1145/3429360.3468180
sha: 7d5778bbaa8331fee12d6ee5607169d357c51fb7
doc_id: 265108
cord_uid: g64c0vzi

Technology based screentime, the time an individual spends engaging with their computer or cell phone, has increased exponentially over the past decade, but perhaps most alarmingly amidst the COVID-19 pandemic. Although many software based interventions exist to reduce screentime, users report a variety of issues relating to the timing of the intervention, the strictness of the tool, and its ability to encourage organic, long-term habit formation. We develop guidelines for the design of behaviour intervention software by conducting a survey to investigate three research questions and further inform the mechanisms of computer-related behaviour change applications. RQ1: What do people want to change and why/how? RQ2: What applications do people use or have used, why do they work or not, and what additional support is desired? RQ3: What are helpful/unhelpful computer breaks and why? Our survey had 68 participants and three key findings. First, time management is a primary concern, but emotional and physical side-effects are equally important. Second, site blockers, self-trackers, and timers are commonly used, but they are ineffective as they are easy-to-ignore and not personalized. Third, away-from-computer breaks, especially involving physical activity, are helpful, whereas on-screen breaks are unhelpful, especially when they are long, because they are not refreshing. We recommend personalized and closed-loop computer-usage behaviour change support and especially encouraging off-the-computer screentime breaks.

Technology has become an intrinsic part of our lives, but excessive technology use is also connected to several physical and psychological problems [11, 18, 30, 31] . Knowledge workers heavily rely on computers [17] and recently, there has been a significant increase in computer usage, especially during the COVID-19 pandemic [6] .

There are several scales for accessing computer usage, e.g., computer use scale [27] , attitudes toward computer usage scale [28] , and compulsive internet use scale [24] . Research has been done to understand as well as modify computer and phone usage, e.g., encourage physical activity [3] , and enable self-tracking [29] and better focus [2, 23] . There has also been research on identifying user needs, e.g., self-monitoring [25] and productivity needs [10, 17] , and what people consider to be work-breaks [9] and what breaks are helpful for productivity [9] .

We investigate 3 research questions to further inform the design of computer-usage behavior change applications, including applications for encouraging computer-related breaks. RQ1. Computer-related Behavior Change Needs and Desired Changes: What what do people want to change about their computer usage and why/how? RQ2.

Currently-used Computer-usage Behavior Change Tools and Further Needs: What techniques or applications do people already use or have used for computer-related behavior change and why do they work or not work, and what kind of additional support do people require? RQ3. Computer-related Helpful and Unhelpful Breaks: What are helpful/unhelpful computer breaks and why are they are helpful/unhelpful?

Research has focused on both understanding and modifying computer and phone usage as well as inferring user states and contexts based on their phone and computer usage, especially for delivering interventions. While previous work focused on specific user needs, e.g., promoting productivity, physical activity, self-tracking, and focus, our work sheds light on the overall computer-usage behavior change needs of the users and highlights what applications and experiences are helpful/unhelpful for the users and why. Our related work is as follows.

There are different scales for accessing computer usage, e.g., computer use scale [27] , attitudes toward computer usage scale [28] , and compulsive internet use scale [24] . Researchers have also used in situ studies to investigate computerusage behavior change needs, e.g., self-monitoring [25] and break prompts that discourage sedentary behavior [20] .

Researchers have also used diary studies and experience sampling to study more holistic scenarios, e.g., combining classic productivity with well-being [10] , understanding personal productivity beyond work-related productivity [17] , and understanding what makes smartphone use meaningful or meaningless [19] . Epstein et al., in particular, conducted a survey to identify what types of breaks, e.g., digital and biological breaks, people consider as breaks from work and what are the desirable qualities of a break, e.g., refreshing, relaxing [9] . Epstein et al.'s diary study focused on helpful breaks for productivity but not necessarily helpful and unhelpful breaks for overall user needs, e.g., physical and emotional health [9] . Our work aims extend this work by surveying the overall user needs, identifying what support works and does not work for users, and which computer breaks are helpful/unhelpful for overall user needs.

There have been several applications to help users monitor and manage their phone and computer usage. While some applications enable passive self-tracking [1, 12, 29] , others employ active interventions, e.g., for self-control on Facebook [21] , for promoting mobility during work-breaks [3] , for regulating phone usage [15, 16, 26] , and for blocking distractions to improve workplace focus and productivity [23] . Researchers have also studied different design choices, e.g., comparing goal-prompt versus removing newsfeed on Facebook [21] , using physiological and location sensing for mobility prompts [3] , comparing a point-of-choice prompt with an always-on progress bar to change sedentary behavior [32] , giving feedback on interruption durations to discourages distractions and interruptions [2] , and using lockout mechanisms [15, 16] or even nudge-like vibrations [26] for regulating phone usage. Researchers have also investigated individual differences in the effects of blocking workplace distractions [22] . However, researchers focus on specific needs, e.g., increasing productivity and reducing distractions or sedentary behaviors, not on overall user needs.

Studies have monitored phone and computer usage, even combined with physiological data, to not only automatically recognize breaks and work activities [8] , but also to model opportune moments for transitions and breaks at work for optimizing happiness and productivity [14] . There is also research to infer opportune moments for well-being messages on mobile phones, e.g., interventions for attention management [4] and for discouraging sedentary behavior [5] . We focus on surveying the helpful/unhelpful breaks and support needs of users to further define the design of computer-usage behavior change interventions.

We conducted an anonymous survey and recruited the participants using convenience and snowball sampling. We shared the survey via department email lists and social media, inviting the participants to share their 'computer-usage patterns and behavior change needs'. There were no explicit inclusion or exclusion criteria for the participants and the participants did not receive any compensation for the survey.

Participants We had 68 participants (35 males, 33 females; = 32.9 years, = 14.8 years; 28 students, 39 full-time workers, 1 retired) from 9 countries -35 from the United States (6 different states), 27 from Malaysia, 24 from the United Kingdom, 4 from Pakistan, 2 from Canada, 2 from India, 1 from France, 2 from Singapore, and 1 from Germany.

Survey Questions: We created our own survey since there was no preexisting survey to investigate our three research questions. We started with Likert scale questions to minimally survey the overall computer usage patterns (Q1-2) and broad problems categories (Q3) of the participants. We did not include full standardized computer usage surveys like CUS to keep our survey short. We then included open-ended questions (Q4-9) to survey the diverse and detailed experiences of our participants for each of our research questions -RQ1: Computer-usage behavior change needs (Q4) and specific desired changes (Q5); RQ2: Currently-used computer-usage behavior change applications and if and why they work or do not work (Q6), and further-desired support (Q7); RQ3: User experiences with helpful computer breaks and why they are helpful (Q8), and similarly for unhelpful breaks (Q9). We iteratively developed the survey questions via peer review and expert review (5 experts and 5 peers) to ensure the validity and reliability of our questions. We also did 5 pilot surveys to further check validity and reliability. All survey questions are in Table 1 .

Data Analysis: For each of the open-ended questions (Q4-9), three researchers independently coded the responses and then collectively performed a thematic analysis of the responses. We performed inductive analysis and the 3 researchers iterated on the codes, themes, and categories for each question before finalizing them. We share the coded responses, themes, and also the top 50 words in each of the responses (excluding words repeated from the question).

We summarize below the results from our survey below for each of the research questions: RQ1. Computer-related Problems and Behavior Change Needs (Q4-5); RQ2. Currently-used Computer-usage Behavior Change Tools and Further Needs (Q6-7); RQ3. Computer-related Helpful and Unhelpful Breaks (Q8-9). Also, we summarize the computer-usage patterns (Q1-Q2) and overall problems (Q3) below. Figure 2 shows the coded responses for desired changes (Q4) and reasons for activity changes (Q5). Figure 4 shows the top 50 words in the responses to Q4 (left) and Q5 (right). 

Q8. Helpful Breaks and Reasons: We divided the responses into 16 categories. Some participants mentioned multiple breaks or one break that fell into multiple categories, e.g., housework also involved physical activity. The most common category of helpful breaks was physical activity (25), followed by housework (9) , time in nature/outside (7), food/water breaks (7), eye exercises (3), family/friends time (3), watching videos (3), doing art (3), etc. Overall, most of the helpful breaks were away from the computer and the breaks were diverse, e.g., 'Nature/outside' was gardening for some and 

Previous computer-related behavior change work has focused on specific goals, e.g., improving productivity, focus and physical activity. We conducted a study to identify the overall user needs. We discuss the key findings, limitations, and recommendations of our computer-usage behavior change needs user survey below. 

Most people spent between 5-10 hours on their computer (Q1, Q2), and time management, emotional problems, and physical discomfort are key concerns for people with respect to their computer usage (Q3). We had three key findings from our survey. First, the participants wanted to reduce time drain, social media usage, entertainment, and physical side-effects (Q4) to better manage their time, emotions, and physical health, and also do less addictive scrolling and 'phone checking' (Q5). Users needs were, thus, have diverse and intertwined needs as they want to reduce their time drain, social media usage, and physical discomfort (Q4) to better manage their time and physical and emotional health (Q5). Second, people use site blockers, time management, and self-tracking applications, but many do not work well, especially because they are easy to ignore and are not designed for different user needs in different contexts (Q6). Thus, users want personalized interventions (e.g., something positive) in personalized contexts (e.g., selective blockers) and

with personalized self-tracking insights (Q7). Third, away-from-screen breaks were usually helpful, especially when they were under 1 hour and involved physical or mental breaks (Q8), whereas on-screen breaks were not helpful and even tended to leave people exhausted and demotivated to work, especially when they were longer than 1 hour (Q9).

We highlight three limitations of our work. First, we conducted our survey during the COVID 2019 pandemic and the computer-usage behavior change needs may be different than 'normal' times since most work was done virtually.

However, given the increasing reliance on technology, our survey highlights the current and future needs for computerusage behavior change. Second, our survey shows diverse participant needs and responses and does not explore one specific area, e.g., overcoming distractions or sedentary behavior. However, it is important to highlight and harness the diverse user needs and experiences to support real-life user needs. Third, we focus on user needs, but the user may not know what is best for them, e.g., user's perceived efficacy may be different from actual efficacy (Q6), and user reports on helpful/unhelpful breaks (Q8, 9) may be biased memories. Thus, it is important to be mindful of user needs and experiences, but then also test them via in situ and longitudinal studies. Overall, our survey is the first of its kind and further surveys and studies may be needed to validate our findings in different settings and with different users.

We have three recommendations. First, even though computer-related behavior change has focused on disjoint goals like productivity, focus, and physical activity, computer-usage behavior change needs are diverse and interconnectedreducing distractions is connected to better time management and productivity, and even physical activity breaks are helpful for "better" computer usage (Q8). The three broad categories of user needs are time management, emotional wellbeing, and physical health, and it may help to provide holistic and personalizable support for user's diverse and interconnected user needs. Second, computer-usage behavior change support can be easy-to-ignore and users want personalized and closed-loop support. We recommend closed-loop behavior change support using reinforcement learning to monitor each user and provide better personalized and context-aware support. Third, off-the-screen breaks are more helpful than on-the-screen breaks. Instead of only restricting on-the-screen breaks, especially since restrictions can be stressful [23] , it might help to replace on-the-screen draining breaks with off-the-screen refreshing breaks.

Also, compulsive technology use is a problem [7] and technology use has a cognitive cost [13] . Thus, we recommend encouraging off-the-screen breaks to curb on-the-screen distractions and boost on-the-screen productivity.

Previous research on computer-related behavior change has focused on goals like productivity, focus, or physical activity, Our findings show that user needs for computer-related behavior change are diverse and interconnected, e.g., the need for better time management is connected to better productivity and focus and even better physical activity as physical movement enables "better" overall computer usage experience. Many computer-usage-related behavior change applications do not work for the users as they are easy-to-ignore and also not personalized for users. Finally,

off-the-screen breaks are, in general, helpful whereas on-the-screen breaks are not helpful.

Thus, user needs are interconnected and users need personalized, holistic, and closed-loop support. One way to offer personalized and closed-loop support is to use reinforcement learning to learn the best context-aware interventions for each user. Also, off-the-computer breaks may be more helpful than on-the-computer breaks. We believe that our findings will inform the design of future behavior change applications for computer usage.

AppInsight: what have I been doing

TimeToFocus: Feedback on Interruption Durations Discourages Distractions and Shortens Interruptions

BreakSense: Combining physiological and location sensing to promote mobility during work-breaks

A Multi-perspective Analysis of Social Context and Personal Factors in Office Settings for the Design of an Effective Mobile Notification System

Multi-Stage Receptivity Model for Mobile Just-In-Time Health Intervention

Coronavirus: Impact on Online Usage in the US-Statistics & Facts. Hamburg: Statista

Compulsive technology use: Compulsive use of mobile applications

A Multi-Sensor Approach to Automatically Recognize Breaks and Work Activities of Knowledge Workers in Academia

Taking 5: Work-breaks, productivity, and opportunities for personal informatics for knowledge workers

Is Your Time Well Spent? Reflecting on Knowledge Work More Holistically

Failure to connect: How computers affect our children's minds-for better and worse

ScreenTrack: Using Visual History for Self-tracking Computer Activities and Retrieving Working Context

Reach for your cell phone at your own risk: The cognitive costs of media choice for breaks

Optimizing for Happiness and Productivity: Modeling Opportune Moments for Transitions and Breaks at Work

GoalKeeper: Exploring Interaction Lockout Mechanisms for Regulating Smartphone Use

LocknType: Lockout Task Intervention for Discouraging Smartphone App Use

Understanding Personal Productivity: How Knowledge Workers Define, Evaluate, and Reflect on Their Productivity

What Makes Smartphone Use Meaningful or Meaningless?

Time for Break: Understanding Information Workers' Sedentary Behavior Through a Break Prompting System

Max Van Kleek, and Nigel Shadbolt. 2020. 'I Just Want to Hack Myself to Not Get Distracted' Evaluating Design Interventions for Self-Control on Facebook

Effects of individual differences in blocking workplace distractions

How blocking distractions affects workplace focus and productivity

The compulsive internet use scale (CIUS): some psychometric properties

Design Recommendations for Self-Monitoring in the Workplace: Studies in Software Development

Good vibrations: can a digital nudge reduce digital overload

PART I: The Computer use Scale: Four Dimensions of how People use Computers

The development of the attitudes toward computer usage scale

Personal Tracking of Screen Time on Digital Devices

Smartphone addiction and its relation to musculoskeletal pain in Egyptian physical therapy students

Smartphone usage and increased risk of mobile phone addiction: A concurrent study

The Point-of-Choice Prompt or the Always-On Progress Bar? A Pilot Study of Reminders for Prolonged Sedentary Behavior Change