201120 Yunha Han

LocknType: Lockout Task Intervention
for Discouraging Smartphone APP Use
Jaejeung Kim, Joonyoung Park, Hyunsee Lee, Minsam Ko, Uichin Lee
CHI 2019 Paper
Slides by Yunha Han

• Overview of the paper
• Related Work
• Lockout Task Intervention Design
• Preliminary Study
• Main Study
• Discussion
• Limitation and Future Work

Overview of the paper
• Motivation of the study
 Negative effects of Smartphones
• Productivity, Safety, Physical/Mental health
-> various strategies to regulate usage, needs for supporting tools
• Goal of the study
 Investigate how a lockout task with varying workloads influence a user’s decision
making.
• proactive intervention that requests users to perform a simple lockout task (typing a fixed length
number) whenever a target app is launched
 Identify the key determinants of a decision making process (in perspectives of app
use) in various contexts in the wild
 Conduct in-the-wild controlled experiment with 40 participants for 3 weeks
• 3 task workloads : 30-digit-input, 10-digit-input, press OK Button
• 3 app categories : web-browser, social media, entertainment
1

Related Work
• Supporting intervention tools
 (Indirect) :
• Usage tracking and visualization -> encourage mindfulness [1]
• Use of social learning and competition [2, 3]
 (Direct) :
• Enabling a blocking mode [4]
• Creating inconvenience by delaying user interaction [5]
• Generating irritative vibration for overuse limitation [6]
• Inserting a mandatory cognitive task before app use [7]
• Proactively blocking in predefined context [8]
2
[1] Heyoung Lee, Heejune Ahn, Samwook Choi, and Wanbok Choi. 2014. The SAMS: Smartphone addiction management system and verification. Journal of medical systems 38, 1 (2014), 1.
[2] Minsam Ko, Seungwoo Choi, Subin Yang, Joonwon Lee, and Uichin Lee. 2015. FamiLync: facilitating participatory parental mediation of adolescents’ smartphone use. In Proceedings of the 2015 ACM International Joint Conference on Pervasive
and Ubiquitous Computing. ACM, 867–878
[3] Minsam Ko, Seungwoo Choi, Koji Yatani, and Uichin Lee. 2016. Lock n’LoL: group-based limiting assistance app to mitigate smartphone distractions in group activities. In Proceedings of the 2016 CHI Conference on Human Factors in Computing
Systems. ACM, 998–1010.
[4] Jaejeung Kim, Chiwoo Cho, and Uichin Lee. 2017. Technology Supported Behavior Restriction for Mitigating Self-Interruptions in Multidevice Environments. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
1, 3 (2017), 64.
[5] Anna L Cox, Sandy JJ Gould, Marta E Cecchinato, Ioanna Iacovides, and Ian Renfree. 2016. Design frictions for mindful interactions: The case for microboundaries. In Proceedings of the 2016 CHI Conference Extended Abstracts on Human Factors
in Computing Systems. ACM, 1389–1397.
[6] Fabian Okeke, Michael Sobolev, Nicola Dell, and Deborah Estrin. 2018. Good Vibrations: Can a Digital Nudge Reduce Digital Overload? (2018).
[7] Joonyoung Park, Jin Yong Sim, Jaejeung Kim, Mun Yong Yi, and Uichin Lee. 2018. Interaction Restraint: Enforcing Adaptive Cognitive Tasks to Restrain Problematic User Interaction. In Extended Abstracts of the 2018 CHI Conference on Human
Factors in Computing Systems. ACM, LBW559.
[8] Inyeop Kim, Gyuwon Jung, Hayoung Jung, Minsam Ko, and Uichin Lee. 2017. Let’s FOCUS: Mitigating Mobile Phone Use in College Classrooms. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 1, 3 (2017), 63.

Related Work
• Theoretical Backgrounds
 Uses and Gratification Theory (UGT)
• to understand why and how people actively use specific media(tv, radio, mobile phones, social
media, internet) to meet their needs [1]
 Expectany-Value Theory (EVT)
• people evaluate their interest and attainment by considering utility and cost -> to explain a user’s
median choice [2]
 Theory of Planned Behavior (TPB)
• one’s belief’s to behavior indicates that an individual’s perceived ease or difficulty of performing a
behavior influences behavior intention [3]
 Social cognitive theory of self-regulation
• Self-regulation has three sub-process : (a) self observation (monitoring one’s behaviors and outcomes), (b)
judgement process (evaluation of observed behaviors as opposed to norms), and (c) self-reaction (adjusting
behaviors based on evaluation results) [4]
• i.e) an individual who spent too much time on smartphone use observes usage amount, judges usage behavior
based on her perceived norm, and utilizes self-control methods to regulate.
3
[1] Adam N Joinson. 2008. Looking at, looking up or keeping up with people?: motives and use of facebook. In Proceedings of the SIGCHI conference on Human Factors in Computing Systems. ACM, 1027–1036.
[2] John D Rayburn and Philip Palmgreen. 1984. Merging uses and gratifications and expectancy-value theory. Communication Research 11, 4 (1984), 537–562.
[3] Icek Ajzen. 1991. The theory of planned behavior. Organizational behavior and human decision processes 50, 2 (1991), 179–211.
[4] Albert Bandura. 1991. Social cognitive theory of self-regulation. Organizational behavior and human decision processes 50, 2 (1991), 248–287

Related Work
• Hypothesis
 Lockout tasks can discourage low benefit gratification seeking behaviors
-> Embed a lockout task into a user’s gratification seeking process
 When a user launches an app for gratification purposes,
a short pause to the instant access making the interaction burdensome (a lockout
task of inputting number) can undermine the desire and intention of using the app
4
When a lockout task is encountered, a user is likely to perform a cost-benefit analysis
involving comparison of he cost (time, effort) of an input task and expected gratification of app use.

Lockout Task Intervention Design
5
• Intervention Principles
 Make an action mandatory to proceed
 Dissociate gratification seeking thoughts via lockout tasks
• Four design dimensions
 Intervention Timing : Lockout tasks are integrated at the time of app launch
 Task Mutability (whether the initially designed rule can be changed or not):
• the scope of target lockout apps -> not allow to change the target app after once selecting
 Task Type and Workload :
• Embedding a “task” to the lockout to create a certain level of cost or workload to the app use
behavior
• Number input tasks are used (to minimize the confounding effects that come from individual
differences in competence and familiarity)
 Target Scope : target an app-level intervention, to discourage particular apps that
could be of negative influence on the user

6
Preliminary Study
• A pilot study to determine ..
 appropriate lockout task and workloads to be assigned to the experimental conditions
 Scope of lockout target apps
 Potential workarounds that may compromise the internal validity of experimental design
• 10 participants (2 female, 4 male, mean age=28.75, sd=4.67)
• Lockout task workload
 0(just press ok button), 10, 30-digit input – randomize workload selection
• Scope of lockout target
 Include counter-productive app (i.e., social media, entertainment etc.)
 Exclude instant messengers and emails
• primarily used for work and communication purposes
• no perceived need for an intervention

Main Study
7
• Goal of the study
 Understand if the insertion of a lockout task at the time of app launch can
discourage app usage
 Explore the determinants that influence such non-use decisions
 Analyze the follow-up behaviors after the use/non-use choices were made
• Participants
 40 participants (18 female, Mean Age=23, SD=3.09) recruited from an online
university community
 Android users who have the intention to reduce their smartphone use

Main Study
8
• Procedure
 1 week of baseline data collection : participants use their smartphone as usual
• At the end of the week, analyzing each participant’s app use behaviors based on three
categories (web browsers, social media, and entertainment) -> selecting the target app
 Following 2 weeks of intervention : deploy lockout task intervention app
• Each user typically received an average of 8.9 intervention target apps (SD=2.4) to register
• After the third week, in-depth interview with 31 randomly selected participants
 In-depth interview
• Q1. the reason behind their decisions for use or non-use on the encountering of the lockout task
• Q2. the follow-up behaviors after the being locked out of the intervention target apps including
attempts of workarounds

9
Results
• Effectiveness of LT Intervention
 LT workload
• Participants experienced three different types of LT workloads in a random manner
• Assess the LT workload using three measures of NASA-TLX, completion time, and the initial
success rate (the first submission of an LT without any error)
ANOVA result confirmed a statistically
significant difference among these three LTS The initial success rate of LT0 was always 100%

10
Results
 Discourage Rate of App Use
• Each user encountered an average of 657 LTs during the two-week intervention period
A two-way ANOVA test shows that
- There was a statistically significant main effect of the
LT workload (F=119.34, P<.001)
- There was a significant main effect of app category
(F=3.72, p=0.025)
- Browser-Social media Pair (p=.041), Browser-Entertainment Pair (p=.010)
- No interaction effects were found between two
variables
Pairwise post-hoc testing using Fisher’s LSD shows that
all LT workload pairs were significantly different (p<.001)

11
Results
 Post-task Usage Behavior : Users chose the following options ..
• Turning off the device (M=25.5%, SD=11.5%)
• Went back to what they have been doing
• Found a non-device activity (e.g., face-to-face chatting with friends)
• Using a non-lockout app (M=50.4%, SD= 16.9%)
• Communication apps (e.g., KakaoTalk) : 55%, Productivity (e.g., calendar) : 13%, Photo/Camera : 7.5%, …
• Using another lockout app (M=24.1%, SD=12.4%)
• Half cases transitioned to the browsers, 25% to the entertainment, 25% to the social media
 Usage Time and Frequency (one sample T-tests with null hypotheses of no change)
• Frequency ratio of LT intervened apps was significantly reduce; close to 50% compared to baseline
-> LT’s usefulness in mitigating frequent use
• No significant difference in the time ratio of LT intervened apps
• LT intervention only helped users to better manage interruptions or the frequent urge to use LT apps
(not in using time)

12
Results
• Determinants of Use/Non-use

13
Results
 User States (at the point of encountering the LT)
• Time availability
• “With only 3 minutes left prior to class, I decided not to use a smartphone, because 3 minutes is not sufficient enough for
desired tasks.” (P3)
• Self-regulation (willingness and mindfulness)
• “I think I just tried not to use my phone and pulled myself together when I grabbed the phone unconsciously. At the
moment when the 10 and 30 digit input task was shown, I was like, ‘Woah, what was I thinking? I don’t need this right
now.’ ” (P4)
• Physical/mental conditions
• “I used my smartphone when I was feeling down. You know, just getting a little comfort from surfing social media and
feeling like you’re connected to this world. So, yes, when I felt depressed, I would unlock my phone anyway even if the 30
digit input was given.” (P1)
• Subjective social norm
• “My girlfriend hates it when I look at my phone when we’re talking to each other. I used to do that quite often,
but since the input task was given and I know it is rude to type it in her presence, I simply focused more on our
conversation instead of staring at my phone.” (P14)

14
Results
 LT Workload Context
• If the time required to perform an LT task was relatively longer than the time a user wishes to use an
app, they tended to choose no to use the app.
• LT30 -> a considerable cognitive burden -> discourage attempts to use an app
• “0 was good, 10 was okay. ... I didn’t really have any thought on it. But 30 was a bit frustrating. ... After all the effort I’ve
put into this task and I still got it wrong! I’m not doing this.” (P12)
• LT30 typically takes less than 20 seconds, whereas the participants perceived that the overall effort
spent for LT 30 was comparable with that of app use that took a much loner time than LT 30
• “Once I got a 10-minute break and 30 digit input task appeared on my game screen. I just gave up. I can’t spend time and
energy on a 30 digit input for a 10-minute break. I thought it was inefficient.” (P16)
 Task Context
• The degree of urgency and importance at a given circumstance affected users’ decision.
• “I really needed to search for the definition of a word to keep up with the lecture. I had to unlock low to high number
input screens every time.” (P10)
• The participants considered whether alternative means of achieving use goals available
• “When my Naver (browser) was blocked with 30 digit input, I tried to access Samsung Internet and Chrome on rotation to
avoid longest digit input.” (P22)

15
Discussion
• Lockout Task as a Behavioral Inhibitor
 Lockout task intervention is an effective tool for discouraging the app use
 The light and gentle interventions (LT0) are effective on the discouragement of app use
(13.1%)
 A slightly higher interaction cost with LT10 double the effectiveness (27.4%)
 Higher cost with LT30 nearly quadruple the effectiveness (47.5%)
• Balancing the Cost-Benefit
 The result of thematic analysis on determinants for decision making indicates that not
only the users’ willingness to reduce their use, but also the users’ intention and
contextual factors also greatly influence.
 Instead of completely blocking use, introduced ‘proactive lockout tasks’
 The contextual norm (e.g., smartphone use during class, family meal) can be considered
in adaptively controlling the lockout task workload for a more effective outcome

16
Discussion
• Follow-up Behavior Guidance
 The thematic analysis of follow-up behaviors revealed that most of the discouraged
cases were followed by positive behaviors
 Open design space for combining intervention approach with a positive follow-up
behavior guidance

17
Limitation and Future Work
• Not consider the user’s positive/negative intention of app use
 Designated the intervention target apps rather than leaving it for the participants to
voluntarily choose -> Increasing the rate of false-positive lockouts
 Focused on understanding both positive and negative lockout experiences despite the
app use intention was a positive or negative one
• The intervention scape (or LT APP Categories) is limited
 Future work can consider other diverse apps with different characteristics
 Future work can consider what problematic usage behaviors exist, and how LT can be
designed to address such problems
• Recruited the participants who at least have considered or tried to reduce
their smartphone use
 Individual differences in the degree of willingness or reduce usage, self-regulatory
capabilities

201120 Yunha Han

More Related Content

Similar to 201120 Yunha Han (20)

More from ivaderivader (20)

Recently uploaded (20)

201120 Yunha Han