This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mhealth and uhealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.
There has been a sharp increase in the number of pedestrians injured while using a mobile phone, but little research has been conducted to explain how and why people use mobile devices while walking. Therefore, we conducted a survey study to explicate the motivations of mobile phone use while walking
The purpose of this study was to identify the critical predictors of behavioral intention to play a popular mobile game,
Participants were recruited from a university (study 1; N=262) and Amazon Mechanical Turk (MTurk) (study 2; N=197) in the United States. Participants completed a Web-based questionnaire, which included measures of attitude, subjective norms, perceived behavioral control (PBC), automaticity, immersion, and enjoyment. Participants also answered questions regarding demographic items.
Hierarchical regression analyses were conducted to examine hypotheses. The model we tested explained about 41% (study 1) and 63% (study 2) of people’s intention to play
Findings from this study demonstrated the utility of TPB in predicting a new behavioral domain—mobile use while walking. To sum up, younger users who are habitual, impulsive, and less immersed players are more likely to intend to play a mobile game while walking.
Despite the efficiency and convenience of communication technology, the increased use of mobile media technology produces alarming safety issues in society, such as mobile phone use while driving [
Despite the prevalence and popularity of mobile device use in everyday life across all age groups, there has been little interest in exploring and understanding what motivates people to use mobile devices, such as a mobile phone, while walking. This study attempts to fill in this gap in prior literature in the psychology of mobile device use. The findings from this study will provide a better understanding of people’s mobile device use in everyday life, which may help health communication researchers and practitioners to create interventions targeting risky behaviors related to mobile device use. The next section provides a short review of mobile phone use literature and the theory of planned behavior (TPB) literature.
There is extensive research examining the effects of mobile phone use on distracted driving [
The pedestrian safety issue, playing augmented reality mobile games such as
Although the potential health benefits of playing
This study uses TPB to find salient predictors of people’s intentions to play a mobile game, in particular
According to TRA, a person’s behavioral intention can be predicted from one’s attitude toward the performance of the behavior and subjective norms. Within the TRA framework, attitude toward a behavior is defined as the extent to which a person evaluates a certain behavior favorably or unfavorably, and subjective norms refer to a person’s perceived social pressure from important others (eg, family members or close friends) about whether one should carry out the behavior or not [
Thus, this study hypothesizes that positive attitudes toward mobile game play (H1), greater perceived subjective norms (H2), and a higher level of PBC (H3) would lead to a strong intention to play mobile games while walking through cognitive evaluation:
Although previous findings have shown the effectiveness of TPB to predict and explain human behavior in many different contexts, researchers argue for the necessity to improve the predictive and explanatory power of the theory with the inclusion of relevant factors such as affective states (ie, anticipated regret), moral norms, or personal norms [
One of the most common criticisms on TPB has been that it only focused on the conscious, reasoned behavioral motivations [
Bayer et al empirically tested the distinct roles of automaticity and immersion in mobile communication [
On the basis of the spectrum, Bayer et al (study 1) explored whether automaticity and immersion independently or simultaneously influence mobile use behaviors such as texting (ie, texting frequency and affective benefits of texting) [
Given that previous research found both automaticity and immersion as significant predictors of mobile use behavior, this research adds these two different behavioral tendencies to TPB, automaticity and immersion in mobile communication. Previous research recommends to simultaneously incorporate two mobile use related−behavioral orientations [
Another factor relevant to playing a mobile game while walking, in general, is an individual’s feeling of enjoyment [
This paper addresses this emerging health issue related to mobile communication by testing 2 samples. Study 1 used a sample of young college students aged 18 to 34 years. Study 2 addressed the lack of diversity of the sample in study 1 by recruiting a sample of people aged 18 to 65 years from Amazon Mechanical Turk (MTurk). The following section details the methodology for this study.
This study used a convenience sample and Web-based survey to ask participants to report their intention of playing
Participants were recruited online from a nonprobability sample (ie, the general communication pool) at a large southern university in the United States from November 2016 to April 2017. Participants were required to be mobile phone users to participate in the complete survey of the psychological processes. In recognition of their participation, they received extra course credit.
As student convenience samples have limited generalizability, the Web-based sample from MTurk was utilized to provide a more diverse sample for this study [
The same measures were used in both study 1 and study 2. Each measure was checked for inter-item correlations, item contribution to scale reliability, and internal consistency. All measures were 7-point scales except for the automaticity and immersion, which used a 5-point scale anchored by 1 (
Participants’ attitude toward playing
To measure the extent to which participants perceived behavioral expectation from their important people, seven items were used [
The extent to which participants perceived that they had control over playing
Automaticity was defined as a behavioral orientation that occurs without conscious awareness. Four items were used to measure the extent of participants’ automatic behavioral orientations toward playing
Immersion was defined as a behavioral orientation that occurs in a conscious manner. Four items were used to measure the extent of participants’ immersive behavioral orientations toward playing
To assess individuals’ feeling of enjoyment while playing
To measure behavioral intention to play
Means, standard deviations, and Cronbach alphas for study 1 variables.
Study 1 variables | Mean (SD) | Cronbach alpha |
Intention | 3.90 (1.32) | .98 |
Attitude | 3.91 (1.32) | .96 |
Subjective norms | 3.86 (1.04) | .82 |
PBCa (after deleting #3) | 6.12 (1.01) | .71 |
Automaticity | 2.09 (0.96) | .87 |
Immersion | 2.37 (0.99) | .93 |
Enjoyment | 4.84 (1.23) | .92 |
aPBC: perceived behavioral control.
Means, standard deviations, and Cronbach alphas for study 2 variables.
Study 2 variables | Mean (SD) | Cronbach alpha |
Intention | 4.28 (1.98) | .97 |
Attitude | 4.68 (1.43) | .97 |
Subjective norms | 4.36 (1.42) | .91 |
PBCa (after deleting #3) | 6.18 (1.08) | .77 |
Automaticity | 2.54 (1.10) | .89 |
Immersion | 2.92 (1.05) | .91 |
Enjoyment | 5.36 (1.33) | .91 |
aPBC: perceived behavioral control.
Finally, questions regarding demographic information, including age, gender, ethnicity, and mobile phone addiction, were measured.
To test the study hypotheses, hierarchical regression analysis was employed. Before conducting hierarchical regression analyses, the categorical variable was dummy coded. For gender, male was coded as 0 and female was coded as 1. Other continuous variables, except the dependent variable, were mean centered to avoid potential multicollinearity [
Correlation matrix of variables in study 1 (N=262), shown as Pearson correlation coefficient r (
Study 1 variables | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | |
1 | Intention | —a | ||||||||
2 | Attitude | .53 (.001) | — | |||||||
3 | Subjective norms | .50 (.001) | .57 (.001) | — | ||||||
4 | Perceived behavioral control | −.11 (.08) | .09 (.14) | .05 (.41) | — | |||||
5 | Automaticity | .25 (.001) | .06 (.38) | .13 (.03) | −.15 (.02) | — | ||||
6 | Immersion | .13 (.03) | .13 (.04) | .14 (.02) | −.10 (.10) | .60 (.001) | — | |||
7 | Enjoyment | .37 (.001) | .42 (.001) | .38 (.001) | −.02 (.79) | .14 (.03) | .22 (.001) | — | ||
8 | Age | −.10 (.10) | .01 (.84) | .05 (.43) | −.02 (.79) | −.07 (.23) | −.01 (.94) | −.08 (.20) | — | |
9 | Gender | .01 (.91) | −.08 (.23) | .01 (.85) | −.03 (.62) | −.12 (.049) | −.09 (.17) | .12 (.05) | −.04 (.54) | — |
a— signifies the correlation of 1.
Correlation matrix of variables in study 2 (N=179), shown as Pearson correlation coefficient r (
Study 2 variables | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | |
1 | Intention | —a | ||||||||
2 | Attitude | .73 (.001) | — | |||||||
3 | Subjective norms | .65 (.001) | .64 (.001) | — | ||||||
4 | Perceived behavioral control | −.05 (.52) | .11 (.14) | .17 (.02) | — | |||||
5 | Automaticity | .27 (.001) | .18 (.02) | .24 (.001) | −.32 (.001) | — | ||||
6 | Immersion | .41 (.001) | .36 (.001) | .32 (.001) | −.05 (.49) | .63 (.001) | — | |||
7 | Enjoyment | .51 (.001) | .56 (.001) | .42 (.001) | .24 (.001) | .19 (.01) | .43 (.001) | — | ||
8 | Age | −.08 (.27) | −.09 (.23) | −.15 (.049) | .09 (.21) | −.15 (.04) | −.05 (.52) | .02 (.74) | — | |
9 | Gender | −.20 (.008) | −.12 (.12) | −.13 (.09) | .12 (.12) | −.32 (.001) | −.29 (.001) | −.03 (.68) | .08 (.26) | — |
a— signifies the correlation of 1.
A total of 417 participants participated in a Web-based survey. After removing individuals who played other mobile games that were not
A total of 264 participants participated in a Web-based survey. After removing individuals who played other mobile games that were not
The results show that the overall model, including all the predictors, was significant,
Hypotheses 4 to 6 predicted that automaticity, immersion, and enjoyment would be significant predictors of intention to play
The overall model, including all the predictors, was significant,
The same procedure that was used in study 1 was conducted to test the Hypotheses 4 to 6. All three TPB components remained significant even after including the additional three variables (see
Regression results for intention to play a mobile game while walking in study 1 (N=262).
Predictors | Standard error (SE) | Beta | ||||||
Age | −.10 | .06 | −.10 | −1.66 | .10 | −.10 | ||
Gender | .01 | .23 | .003 | 0.05 | .96 | .003 | ||
Attitude | .54 | .08 | .40 | 6.59 | <.001 | .32 | ||
Subjective norms | .49 | .10 | .29 | 4.76 | <.001 | .23 | ||
PBCe | −.27 | .09 | −.16 | −3.15 | .002 | −.15 | ||
Attitude | .51 | .08 | .38 | 6.13 | <.001 | .29 | ||
Subjective norms | .41 | .10 | .24 | 4.06 | <.001 | .19 | ||
PBCc | −.23 | .09 | −.13 | −2.74 | .007 | −.13 | ||
Automaticity | .44 | .11 | .24 | 3.92 | <.001 | .19 | ||
Immersion | −.25 | .11 | −.14 | −2.31 | .02 | −.11 | ||
Enjoyment | .16 | .08 | .11 | 2.04 | .04 | .10 |
a
b
c
d
ePBC: perceived behavioral control.
f
Regression results for intention to play a mobile game while walking in study 2 (N=197).
Predictors | Standard error (SE) | Beta | |||||
Age | −.02 | .02 | −.07 | −0.91 | .36 | −.07 | |
Gender | −.76 | .30 | −.19 | −2.59 | .01 | −.19 | |
Attitude | .73 | .08 | .53 | 8.73 | <.001 | .41 | |
Subjective norms | .46 | .09 | .33 | 5.28 | <.001 | .25 | |
PBCe | −.29 | .09 | −.16 | −3.29 | .001 | −.15 | |
Attitude | .60 | .09 | .44 | 6.56 | <.001 | .30 | |
Subjective norms | .44 | .09 | .32 | 5.06 | <.001 | .23 | |
PBC | −.35 | .10 | −.19 | −3.63 | <.001 | −.17 | |
Automaticity | −.06 | .12 | −.04 | −0.54 | .59 | −.03 | |
Immersion | .14 | .12 | .08 | 1.13 | .26 | .05 | |
Enjoyment | .23 | .09 | .15 | 2.55 | .01 | .12 |
a
b
c
d
ePBC: perceived behavioral control.
f
The three TPB variables (attitudes, subjective norms, and PBC) were significant predictors of intention to play
Compared with the long tradition of research on people’s perception, attitudes, and behaviors affected by newspaper, magazines, radio, television, and the Internet, we have relatively little understanding of the psychological and behavioral impact of mobile media. There has been a drastic increase in pedestrian injuries while using mobile phone, but little research has been conducted to explain individuals’ motivations to use mobile devices while walking. This research identifies the predictors of intention to play a popular mobile game,
Overall, the model we tested explained 41% of people’s intention to play a mobile game while walking in study 1 and 63% of people’s intention to play a mobile game while walking in study 2, with attitude toward playing a mobile game while walking emerging as the strongest predictor, followed by subjective norms. Both in study 1 and study 2, the three TPB components significantly predicted one’s intention to play a mobile game while walking. The three TPB components explained 37.6% in study 1 and 58% in study 2 of behavioral intention to play a mobile game in this study. The results indicated that attitude and subjective norms were positive significant predictors of intention to play a mobile game while walking, whereas PBC was a negative significant predictor of intention to play a mobile game while walking. The more they thought that playing the mobile game while walking was positive and beneficial, the greater their intention to play was; the more they believed that others would like them to play the mobile game while walking, the greater their intention to play was. However, the less they felt that they had control over playing it while walking, the greater their intention to play was.
The results demonstrated the utility of TPB in the context of mobile-related health behavior. TPB has been applied to many health-related behaviors including behaviors that cause public safety issues such as distracted driving [
Especially, the significant effect of normative beliefs is thought-provoking in that previous studies pointed out that mundane tasks such as road crossing were less likely to be influenced by normative beliefs [
Contrary to the hypothesis 3, there was a negative association between individuals’ PBC and behavioral intention to play the game. In fact, previous research pointed out that a lesser degree of PBC leads to greater behavioral intention [
Further supporting this point, the feeling of enjoyment was a positive significant predictor of intention to play a mobile game in both study 1 and study 2. People who felt greater enjoyment were more likely to intend to play the game while walking than those who felt less enjoyment. This finding is consistent with prior literature on the effect of enjoyment on intention to play a Web-based game [
This study also found that automaticity and immersion were significant predictors of playing the mobile game in study 1. The automaticity and immersion factors increased by 4.4% of the amount of explained variances in intention to play a mobile game in addition to enjoyment factor in study 1. As college students played the mobile game more habitually and impulsively, they were more likely to play it while walking. This implies that people who have a tendency to play mobile game without intending to do so are more susceptible to perform such behaviors, which is consistent with prior literature’s findings on habitual mobile media use [
Our correlation results reported in
In fact, it has to be noted that individuals’ levels of automaticity and immersion were significantly, and positively, correlated with each other in both studies. Bayer et al also found that these two tendencies are highly correlated in texting behavior—individuals who habitually text also do so with more immersion [
Yet, these behavioral tendencies were not significant predictors of the intention of playing a mobile game while walking in study 2. In study 2, enjoyment was the only additional predictor of intention to play apart from the three components of TPB. It should be noted that there were differences in terms of sample characteristics between study 1 and study 2. The participants in study 1 were recruited from a university, whereas study 2 participants were recruited through MTurk, the crowdsourcing website that is known to be more representative of the US population. In addition, a post-hoc analysis using an analysis of variance (ANOVA) revealed that there was a significant difference in the degree of mobile phone addiction between study 1 (mean 5.07 [SD 0.99]) and study 2 (mean 4.74 [SD 1.26]). Participants in study 1 reported greater levels of mobile phone addiction (
For these college students, their intention to play a mobile game can be significantly explained by their behavioral tendencies, because playing a mobile game is more than mere enjoyment to them. For addicted users, their habitual and impulsive tendency of playing a mobile game while walking would be an additional motivational factor that leads to greater intention to play. In addition, younger users’ attention is more likely hijacked by immersive media such as video games [
This finding suggests that playing a mobile game such as
In addition, there was a considerable difference in the amount of explained variances in intention: study 1 explained 41%, whereas study 2 explained 63% of variance in their intention to play. Our model focused on attitudes, subjective norms, and PBC as main predictors, which did not address individuals’ digital media use patterns in general, including the level of addiction discussed above. College students in study 1 may have already developed their own digital media preference and could have been more experienced in mobile game play, whereas older users in study 2 may have been less influenced by these factors. Future research may benefit from identifying more media use variables for predicting college students’ mobile game play while walking.
The studies’ findings have implications for health promotion practice and policy. First, health practitioners and designers should incorporate normative information in their campaign messages to amend people’s false beliefs and optimistic biases, and thus change such habitual and addictive behaviors in a desirable way. Second, inattention is one of prevalent causal factors of road incidents [
This study investigated playing a mobile game as the only target behavior. However, it seems significant to examine other mobile usage behaviors (eg, social media use such as Snapchat, listening to music, or taking pictures or videos while walking), given that they may trigger different types of safety issues from playing a mobile game. In addition, it is still unclear how different predictors of TPB would be applied to a different set of mobile behaviors. For instance, subjective norms were not a significant predictor for crossing streets while listening to music or texting [
By showing the significant effect of individuals’ consciousness tendency on behavioral intention to play mobile games while walking, this study suggests the need for future studies on habitual or addictive mobile media use. Individuals who habitually use mobile media while walking, without meaning to do it, might be more likely to experience serious health threats such as accidents. Therefore, future studies ought to investigate some predictors of the unconscious media use, including personality and situational factors, and further analyze how these factors are associated with mobile game play in unsafe settings.
Our sample size was not big, limiting the generalizability of our results. After selecting only those who played
Survey questionnaire.
analysis of variance
Amazon Mechanical Turk
perceived behavioral control
standard deviation
standard error
theory of reasoned action
theory of planned behavior
None declared.