Skip to main content

Investigation of Smartphone Use While Walking and Its Influences on One’s Behavior Among Pedestrians in Taiwan

Part of the Communications in Computer and Information Science book series (CCIS,volume 714)


While bringing about convenience, using smartphones while walking may distract the users, leading to potential hazards for both the user him/herself and other road users. In recent years, the number of accidents resulting from smartphone use while walking keeps growing. It hence becomes urgent to develop countermeasures to solve this problem. Therefore, this study aims to investigate the current status of smartphone use while walking in Taiwan, as well as analyzing its influences on the user’s behavior.

On the one hand, a structured questionnaire was administered to 100 participants ranging from 20 to 64 years old, in which one’s experience of and attitude toward smartphone use while walking being surveyed in detail. On the other hand, behavioral data were collected and analyzed among three male and three female adults in their 20 s by an eye tracking system and a motion capture system. The performance and strategies of foot-eye coordination were compared among conditions including free walking, walking while holding a smartphone, and texting on a smartphone while walking.

Results showed that most people are aware of the risks associated with smartphone use while walking, but they somehow fail to follow the recommendations for not doing so. The findings also suggest that distracted walking caused by smartphone use did slow down the pedestrian, limit his/her visual attention, and impair the stability of walking. Therefore, countermeasures would be necessary to help reduce the associated risks.


  • Smartphones
  • Distracted walking
  • Foot-eye coordination

1 Introduction

According to the Institute for Information Industry (2015), the possession rate of smartphone in Taiwan is now approximately 73.4%. In addition, it seems that smartphones have also gradually become one part of middle-aged and older adults’ daily lives, with the possession rate reaching 26.6%. However, the dependence on smartphones may lead to problems such as phone/Internet addiction, increased stress, poor social relationship, and musculoskeletal disorders (Lanaj et al. 2014; Hassanzadeh and Rezaei, 2011; Eom et al., 2013). Further, as more and more people keep using smartphones while walking, the associated cognitive or physiological changes in one’s behavior can reduce risk awareness of potential hazards.

Fox and Duggan (2012) reported that 53% of smartphone users have the experience of colliding with others due to distracted walking, in which the majority is the young people aging from 18 to 24 years old. In the United States, hospital statistics indicated that the number of patients sent to the ICUs due to phone-use issues reached 1,152 in 2011 (Pew Charitable Trusts 2014). In Japan, there were 122 pedestrians sent to hospitals because of accidents associated with phone use (Tokyo Fire Department 2014). The Road Traffic Authority of Korea found that the growth rate of phone-related emergency services is 1.7 times more than that of the countries in OCED (Molko 2016). The case is similar in Taiwan, but there were very limited studies or surveys addressing related issues.

In addition to the statistics of preferences and accident histories, behavioral changes due to distracted walking while using smartphone are also of great concern. For example, Haga et al. (2015) invited participants to walk clockwisely along a 3 m-by-3 m square, during which a smartphone was used for texting, watching video clips, and playing games. Their observations suggest that distracted walking with smartphones may lead to greater deviations from the walking route and reduced attention to unexpected visual stimuli. Nevertheless, the participant’s gaze behavior was not analyzed, and quantitative data of the lower-limb movements were limited. Further, how auditory attention is affected was also not studied.

Therefore, this study aims to investigate the experience of and attitude toward smartphone use while walking among Taiwanese people, as well as analyzing the effects of such distracted walking on one’s gaze behavior, performance and strategies of foot-eye coordination, and residual capacity of both visual and auditory attention.

2 Investigation of Users’ Preferences

A structured questionnaire covering the demographics, experiences of smartphone use, attitude toward distracted walking with a smartphone, and the acceptance of few countermeasures against smartphone use while walking was administered to 100 participants ranging from 20 to 64 years old and collected in one month. All of them own a smartphone and agreed with the consent statement for the questionnaire.

78% of the participants are under 50 years old (23 are between 20 and 24 years old; 23 are between 25 and 29 years old; 10 are between 30 to 34 years old; 10 are between 35 and 39 years old; 8 are between 40 to 44 years old; 4 are between 45 and 49 years old), while 22% are over 50 years old (10 are between 50 and 54 years old; 11 are between 55 and 59 years old; 1 is between 60 to 64 years old). In addition, there are 57 males and 43 females.

Considering the time spent with the smartphone, 49% of the participants are frequent users who spent more than 4 h per day. In the cases of smartphone use while walking, 78% of the participants have had such experiences. Among these 78 participants who have used smartphones while walking, only 14% of them did this frequently. Further, 36 out of these 78 participants used their smartphones while walking with social media applications (e.g. Facebook, LINE), followed by talking on the phone (14 out of 78), reading text messages (6 out of 78), listening to music (6 out of 78), and so forth. As for those 22 participants who never used smartphones while walking, their reasons include “feeling it’s dangerous” (33%), “avoiding collision with other road users or obstacles” (21%), “avoiding trips or falls” (21%), and so on.

Despite the high percentage of the experience of smartphone use while walking, only 22% of them have hence got involved in accidents such as collision with other road users or obstacles. This is probably one of the major reasons that encouraged them to keep doing this, even though 78 out of the 100 participants disagree with the use of smartphones while walking. From this point of view, people’s attitude toward smartphone use while walking may conflict with how they actually behave, which hinders the promotion of corresponding countermeasures.

Moreover, in the questionnaire, the participants were asked to rate from 1 (strongly unacceptable) to 5 (strongly acceptable) about their attitude toward each of six countermeasures designed for preventing smartphone use while walking. The most acceptable countermeasure (with the average of 3.66 out of 5.00) was the introduction of a transparent screen that allows the user to see through the screen for keeping attention to the information around his/her feet. Following that, the simplified design of user interface (requiring less time and attention) also received a rating than 3.50 (out of 5.00). On the contrary, pop-out windows warning that “stop using smartphones while walking” was least acceptable (with the average of 2.82 out of 5.00), followed by auto-locking of the screen as the users is detected as walking (with the average of 3.15 out of 5.00). In short, people tend to prefer technologies that improve one’s attention and allow smartphone use while walking, rather than those prohibit them from such a multitasking behavior.

3 Behavioral Changes Associated with Distracted Walking

In order to quantify the behavioral changes associated with distracted walking due to smartphone use, six participants (3 males and 3 females) ranging from 22 to 26 years old were recruited to perform simulated tasks in the laboratory environment. The body height is 168.3 ± 1.5 cm for males and 158.7 ± 7.0 for females, whereas the body weight is 62.0 ± 2.0 for males and 46.7 ± 4.2 for females. All of them have the experience of smartphone use while walking and signed the informed consent prior to the experiment.

Each participant was asked to wear a mobile eye-tracking system, the standard suit for motion capture with 37 reflective markers attached, and a pair of standard shoes that fits his/her feet. First, the participant was allowed to get used to walking along the 5.1 m by 3.6 m square, in which the route width was set at 0.3 m. There was a monitor placed 1.6 m away from the end of each longer side (5.1 m), which was prepared for presenting unexpected stimuli (pure colors of red, yellow, or green, lasting for 60 ms) as the measure of the participant’s residual capacity of visual attention when walking on any of the two longer sides. In addition, a wireless speaker was placed at the center of the square, so that the sound recorded in a crowded street (with the loudness of 65 dB) can be played while the participant performs the task. Once the participant is on any of the two shorter sides (3.6 m), a horn (with the loudness of 70 dB) might appear for 0.5 s to test his/her residual capacity of auditory attention. Further, a digital video recorder was placed at the corner opposite to the starting point. The layout is shown in Fig. 1.

Fig. 1.
figure 1

Layout of the simulated route (Color figure online)

After practicing walking along the route shown in Fig. 1 with and without the use of the standard smartphone, there were six sessions assigned for each participant. In each session, the participant needs to walk along the route for four laps in his/her preferred pace while the corresponding tasks are performed. In the first session, the participant was required to keep standing on the starting point and use a social media application to interact with a chatting robot through texting. In the second session, the participant also kept standing on the starting point while playing games on a typing test application. These two sessions were designed as controls to see how the participant’s performance may change as he/she was required to complete the same tasks while walking. In the third session, free walking along the route for four laps was required, as shown in Fig. 2 (left). This is considered as the baseline of one’s behavior under such an environment. In the fourth session, the participant walked along the route while holding a smartphone without really using it for four laps. This is used to clarify whether the behavioral changes are purely due to smartphone operation, or the fact that one is holding it as well (Fig. 2; right). The fifth session is similar with the first session, except that the participant had to walk along the route for four laps. Finally, in the sixth session, the participant did the same smartphone task as in the second session, but he/she was required to walk along the route for four laps. In sessions 3 to 6, the participant’s gaze data and motion capture data were collected throughout the four-lap movement. Meanwhile, the behaviors were videotaped using the digital video recorder, while the operations on the smartphone were recorded using a remote control application. Further, the participant’s responses to the visual/auditory stimuli were collected in terms of response time (from the appearance of stimulus to the beginning of the participant’s verbal reaction) and percentage of correct responses (whether the participant spoke out the correct color or the presentation of the horn).

Fig. 2.
figure 2

Free walking (left) and walking while holding a smartphone (right) (Color figure online)

When there was smartphone operation involved while walking (session 5 and session 6), the participants generally moved the gaze away from the smartphone only as turning at the four corners (Fig. 3; left), whereas the gaze was mostly on the smartphone no matter whether they walked on the longer or shorter sides (Fig. 3; right). In the case of walking on the longer sides, the participants tended to look at a higher level, which seems to be prepared for capturing the upcoming visual stimulus. However, while performing the chatting task, the participant usually spent some time thinking about what to communicate with the chatting robot. So, at these moments, some of the participants moved their gaze away from the smartphone slightly longer, allowing more residual capacity of visual attention.

Fig. 3.
figure 3

Gaze away from (left) and on (right) the smartphone (Color figure online)

As illustrated in Fig. 4, considering the stability of walking (along the center line of the route), conditions while performing the typing task or the chatting task showed greater lateral deviations. This conforms with the assumption that distracted walking may increase the difficulty of walking as the attention is drawn to the tiny screen of the smartphone. In addition, it was found that the lateral deviation of walking while holding a smartphone is quite similar with that of free walking. In other words, it is very likely that gait changes associated with smartphone use while walking were purely due to the decreased attention caused by smartphone operation.

Fig. 4.
figure 4

Lateral deviation from the center line of one participant (Color figure online)

Moreover, the walking speed (measured in terms of task completion time throughout the four laps) was found to be significantly reduced as the smartphone was used while walking (p = 0.049). The mean completion time for four laps was 77.1 s, 79.1 s, 99.1 s, and 103.9 s for holding, free walking, chatting, and typing, respectively. Post-hoc analysis revealed that the participants walked significantly more slowly while performing the typing task than in holding and free walking conditions. Obviously, it is very likely that the high level of the participant’s attention on the smartphone changed his/her gait pattern in terms of both accuracy and efficiency.

Generally speaking, the use of smartphone requires more visual attention than auditory attention. Thus, following the multiple resource model (Wickens, 1980), one should be more sensitive to unexpected auditory stimuli than visual ones. However, surprisingly, the increase of response time to auditory stimulus (p = 0.049) was found to be as significant as that to visual stimulus (p = 0.033), while comparing between conditions with and without smartphone use. In the case of visual stimulus, the mean response time was 0.665 s, 0.715 s, 0.895 s, and 0.969 s for holding, free walking, chatting, and typing, respectively. Post-hoc analysis showed that the response time was significantly longer in chatting and typing tasks than in holding and free walking tasks. As for auditory stimulus, the mean response time was 0.619 s, 0.655 s, 0.808 s, and 0.886 s for holding, free walking, typing, and chatting, respectively. Post-hoc analysis indicated that the response time was significantly longer in the chatting task than in holding and free walking tasks.

Nevertheless, on the other hand, the percentages of correct response to visual stimuli were found to be significantly different among conditions (p = 0.001), whereas there was no difference in the percentage of correct response to auditory stimuli (p = 0.422). Post-hoc analysis showed that there were significant differences for visual stimuli between any pair of conditions, except for holding versus free walking and typing versus chatting. In other words, though it took a longer time to respond to auditory stimuli as one uses smartphone while walking, it doesn’t affect how accurately the response is. As for visual attention, both capacity and accuracy were impaired.

4 Conclusion

In this study, the investigation of the experience of and attitude toward smartphone use while walking was conducted among 100 Taiwanese people ranging from 20 to 64 years old. Statistics showed that most people tend to behave in this way, even though they are aware of the risks associated with it. Hence, countermeasures that both allow people to do so and improve the safety would be required. On the other hand, findings of the experiment suggested that distracted walking due to smartphone use did make the user walk more slowly and less stably, as well as limiting the visual attention. Further analyses on gaze behavior and lower-limb kinematics will be continued, so as to highlight the importance of corresponding countermeasures and to identify reliable measures for testing and evaluation.


Download references

Author information

Authors and Affiliations


Corresponding author

Correspondence to Jun-Ming Lu .

Editor information

Editors and Affiliations

Rights and permissions

Reprints and Permissions

Copyright information

© 2017 Springer International Publishing AG

About this paper

Cite this paper

Lu, JM., Lo, YC. (2017). Investigation of Smartphone Use While Walking and Its Influences on One’s Behavior Among Pedestrians in Taiwan. In: Stephanidis, C. (eds) HCI International 2017 – Posters' Extended Abstracts. HCI 2017. Communications in Computer and Information Science, vol 714. Springer, Cham.

Download citation

  • DOI:

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-319-58752-3

  • Online ISBN: 978-3-319-58753-0

  • eBook Packages: Computer ScienceComputer Science (R0)