Most citizens of developed countries have access to and use cell phones or similar mobile devices to transfer information and data.1–3 Cultural demands require most users to engage in use of mobile technologies to be a part of employment, education, recreation, or social-life.4–6 Additionally, there are socially implied sanctions imposed for people who do not keep up with and use mobile technology frequently.7 Mobile phones have replaced some other devices (e.g., cameras, address books) but may become an object of attachment.8 There may also be adverse health consequences associated with mobile phone use.9 Glucose was elevated in the orbitofrontal cortex of adults on the side of the brain adjacent to a mobile phone receiving a call.10 These neural changes were temperature independent. Glucose was reduced in the temporoparietal junction and anterior temporal lobe of young adults (age 21 to 29) on the same side as mobile phone exposure.11 There is also an extensive body of data utilizing electroencephalographs.12 A meta-analysis identified subtle, but significant, reductions in sperm motility (−8.1%) and viability (−9.1%) among mobile phone users.13 One hour of exposure to a mobile phone by young-adults in their early twenties increased plasma lipid peroxidases and decreased antioxidant enzymes.14 Understanding whether these acute effects increase the susceptibility to certain diseases has been the topic of intense interest.15 Mobile phone use does not appear to increase the risk for glioblastoma,16 and microarray studies have generally been unable to identify a proteomic profile that is beyond what would be expected by chance.17
Of several epidemiological findings in this area, distracted driving is arguably the best established and most consequential risk. There were 3157 documented automobile fatalities that involved distracted driving in the United States in 2016. Of these, more than 400 involved a mobile phone in use at the time of the accident.18 Even these numbers may be underestimates, as they do not take into account how texting at night contributes to drowsy driving,19 which could further increase motor vehicle accidents. Two-thirds of adults (age 18 to 65) in the United States reported talking on their mobile phones, and one-third received or sent a text or email while driving in the past month.20 Even for the minority who do not choose to use mobile phones, proximity to the devices carried by their peers and others can impose risks. Non-users can be the victims of injuries sustained by user inattention.21 A video game for mobile devices resulted in accidents.22 On the other hand, public health investigations have generally found that text-messaging interventions can produce positive outcomes for diabetes self-management, weight loss, smoking cessation, and medication adherence.23,24 A behavioral component to wireless technology use may increase unintentional injury and mental health issues.25 Overuse or misuse of this technology may have parallels to addiction,2,26,27 although further research may be needed to establish this.28 People with addictive behaviors are not always able to make rational decisions regarding that behavior.2,29,30 For some users, an inordinate amount time is spent in non-productive activities that cut into time spent on healthier endeavors.21,31
The millennial generation, born between 1982 and 2002,32,33 may have a special relationship with technology.7,22 This is the first generation to grow up with mobile information-based technology being likened to adequate educational preparation and success as an adult.34 Educational leaders have spent an enormous amount of time, money, and administrative resources continuously upgrading modern education paradigms to produce technologically friendly environments.35,36 Millennials are more likely than prior generations to demand technology and use it in innovative ways to enhance their lives.37,38 However, it may turn out that social activities rather than academic or professional ones are most aided through their use of technology. Millennials, primarily traditional college-age students, spent an average of 14 hours texting, 6.5 hours talking on the phone, and 6.5 hours using social media sites per week, with some sending more than 200 texts in a day.39
The Theory of Planned Behavior has been refined by Icek Ajzen and colleagues over the past four decades.40–42 Behavior is viewed as guided by beliefs about the consequences, the normative expectations of others, and the presence or absence of factors that may impact it. These behavioral beliefs, normative beliefs, and control beliefs contribute to the attitude toward the behavior, the subjective norm, and perceived behavioral control, respectively, which together result in behavioral intention.41 The Ajzen model has been widely utilized in such diverse areas as nutrition,43 exercise,44 and COVID-19 vaccination.45
The purpose of this study was to explore the likelihood of behavior change regarding mobile devices following a brief educational intervention to millennials. This research question was embedded within the Theory of Reasoned Action and Planned Behavior.40–42
The participants were young-adult college students (N = 215, 67.0% female, age = 20.0 + 1.6, min age = 17, max = 28). Exclusion criteria were not being a college student and age > 30.
Research team members held informal meetings to develop reliable quantitative surveys with items based in concepts from the Theory of Reasoned Action.40–42 Drafts of each instrument were piloted using demographically similar non-participant groups.
A comprehensive search of the literature was conducted regarding the potential for negative health effects associated with mobile phone use including sleeplessness, injury and other situations and behaviors deleterious to human health. Using these results,6,10–13,20,21,25,31,46–50 a one-page educational fact sheet was produced to use as an intervention tool. When instruments and tools were satisfactory to the research team, approval was obtained from a university Institutional Review Board.
An experimental study design was employed. It was a one-week pre-post-intervention design completed in 2016, with a subset of participants also completing a 30-day follow-up. Due to the potential benefits of the intervention, delayed intervention was used with the comparison group, assuring the ethical treatment of all participants. The data-collection protocol was determined by the research team, and data-collection packets were provided to each data collector. Due to the public-health risks associated with mobile phone use (e.g., distracted driving), participants were assigned to the intervention or comparison group in an approximately 3 to 1 ratio. Each data collection packet held a pre-coded participant log with pre-determined selected numbers identifying comparison group participants (participant #2, #5, #7, #14, #20, etc.), the informed consent, and 20 pre- and post-intervention coded surveys to provide identification for follow-up.
The one-page printed fact sheet was provided to the intervention group immediately after the pre-intervention survey was completed. The comparison group received it immediately after the second post-intervention survey (i.e., one month later). (See Appendix for the instrument). There were seventeen volunteer data collectors who were trained to gain written informed consent, administer the survey, and provide the fact sheet effectively. Each data collector was charged to recruit twenty willing potential participants who would be assigned to either an intervention or a comparison group. All potential participants were “cold contacts” of the millennial generation (age < 30) attending a public liberal arts institution in the northeastern United States. Recruitment commonly occurred in public areas. No incentives were offered for participation.
As part of the pre-intervention survey, demographic and general mobile phone information (type and basic usage frequency) were collected. The paper survey also included items based in the Theory of Reasoned Action and Planned Behavior, transitioning from values to attitude and intention, such as: “I believe there may be negative health effects related to cell phone use” and “I should change some of my behaviors when it comes to my cell phone use” (Appendix). One week later, the first post-intervention survey was administered to intervention and control groups. A subset of both groups completed the post-intervention survey at one-month after the baseline. The fact sheet was shared with the comparison group members at this juncture after the second post-test data had been obtained.
Descriptive statistics were used to describe the characteristics of both groups. We compared mobile phone use and attitudes and intentions for the two groups by calculating the percentage of agreement with survey items, which was analyzed with a chi-square with Yates correction. Differences between groups on parametric measures were expressed in terms of an independent samples t test and, when significant, Cohen’s D. Variability was expressed as the SD unless noted otherwise. Comparisons between the intervention versus the control group was analyzed in SYSTAT, version 13.1 (Chicago, IL) with α = .05. Internal consistency was determined with Cronbach’s α. A total index of mobile phone attitudes and behaviors was calculated by summing the seven post-test ratings (where 1 = strongly agree to 5 = strongly disagree), with the two negatively worded items reverse scored. Figures were prepared with GraphPad Prism, version 6.07 (La Jolla, CA).
Participant demographics and baseline characteristics are in Table 1. No demographic or mobile phone–related variables differed significantly between groups at baseline. Parametric analysis identified one item, “I sometimes miss things going on around me (i.e., conversations, class lectures, etc.) because I am doing something on my cell phone,” at baseline between the comparison (2.5 ± 0.7) and the intervention (2.8 ± 0.8, t(210) = 2.85, P ≤ .005) but no other differences (P > .18). The internal consistency of the 12-item mobile-phone baseline assessment (Appendix, questions 1–12) was 0.78.
Figure 1 depicts the mean ratings at the one-week post-test. The intervention scored significantly lower (1 = strongly agree to 5 = strongly disagree) on three of the five positively worded items (d = .32, .44, and .35). The fourth item, “I plan to keep my cell phone away from my body more often,” approached common statistical thresholds (t(213) = 1.93, P = .054). The two reverse-worded items showed the opposite pattern with significantly higher ratings (d = .40 and .47). Non-parametric analyses showed the same general pattern with significant differences favoring the intervention on four items (Supplementary Figure 1).
The internal consistency of the 7-item post-intervention measure was 0.82. Therefore, a total index of mobile phone behaviors was created. The Intervention scored lower than the Comparison (Figure 2, t(210) = 3.16, P < .005, d = .50). The effect size was larger in women (d = .57) than men (d = .38) but similar in smart phone (d = .54) and non–smart phone (d = .47) users.
At one month after the baseline, a subset of participants (33.3% of the original comparison group and 39.4% of the intervention group) were reassessed for the second post-test. Completers were 1.5 years younger than non-completers (completers = 19.2 ± 1.5, non-completers = 20.6 ± 1.3, t(213) = 7.41, P < .0005). More completers (79.5%) than non-completers (59.2%, χ2(1) = 8.85, P < .005) viewed their mobile phone as a necessary expense but groups were otherwise indistinguishable (Supplemental Table 1). There were no differences between the intervention and comparison on individual items (t test P > .09) or the total score (t test P = .32). However, as shown in Figure 2, the intervention at one month was still lower than the comparison at one week (t(122) = 2.62, P ≤ .01). Further, the comparison at one month was significantly decreased relative to the comparison at one week (t(82) = 3.61, P ≤ .001). The test-retest correlation between post-tests at one week and one month was limited for all participants (r(84) = −.04) and within each group separately (comparison: r(22) = −.14; intervention: r(60) = .01). The majority of the intervention condition agreed that they learned something new from the fact sheet at the first (87.1%) and second (65.6%) post-tests.
This report makes two contributions to the public health and behavioral addiction fields. First, this controlled study demonstrates that mobile phone use attitudes and behaviors are fluid and perhaps at least transiently malleable in young adults. Second, this investigation provides some preliminary psychometric support for the development of a relatively abbreviated instrument to quantify mobile phone behaviors. The prevalence and use of information technology by millennials in the form of mobile devices is well established.3,9,21 A mounting body of evidence suggests that exposure to electromagnetic radiation emitting devices has detectable, and possibly deleterious, effects on human health and well-being.10,12,13,15,21,47,48,51,52 Currently, the best-established public-health risk associated with mobile phones may be distracted driving.25 Young adults (age ≤ 29) were over-represented fatal crashes involving mobile phone use while driving.18 Sleep is a fundamental neurobiological process, but texting is part of a profile of poor sleep hygiene by many adolescents and young adults.21,50,53 Light-emitting electronic devices can also reduce melatonin.54 Many people, especially those from the millennial generation, find it challenging to comport their lives without continuous wireless access.
To date, little is available in popular culture to alert users of potential risks. Notably, an early association between cell phone use and brain cancer15 did generate some media attention, but this association was not substantiated in subsequent research.55 Only a few public policies curtail or prohibit mobile device use in specific environments (e.g., texting while driving; mobile phone use on an airplane, in a movie theater, during an educational lecture, near a gasoline pump, or while interacting with a health care provider). Risks are commonly associated only with special circumstances, in which adherence and enforcement may be culturally subjective. This may send the message that all other use is safe. The present findings that mobile phone attitudes and behaviors can be modified, at least transiently, may be of particular interest to mental health providers with patients, or their family members, interested in reducing mobile phone use.27 Further, the development of mobile phone interventions may be useful for individuals caught repeatedly violating laws prohibiting texting and driving.
While most time spent on mobile devices by millennials is for social or recreational reasons, they also use their mobile devices as tools for information and navigation throughout their day (e.g., social media, email, clocks/calendars, GPS, audio and video entertainment). The feasibility of curtailing that multifaceted usage has not yet been well established. This investigation identified a small-moderate effect size with a brief-intervention at a short (one-week) interval.
Although not the primary objective, this study also adds to other findings to develop other abbreviated measures of problematic mobile phone use.56–59 A Smartphone Addiction Inventory has been developed for users who can read Mandarin which showed excellent internal consistency and very good two-week test-retest reliability in a primarily male young adult sample.60 The internal consistency of both the baseline and post-test instrument used in this effort was satisfactory. However, this was not the case for the test-retest reliability. However, as only a minority of participants were available for the one-month post-test, any inferences regarding the data at this interval (correlations or means) should be made with caution. It is noteworthy that the scores on two individual items, “I should change some of my behaviors when it comes to my cell phone use” and “I plan to keep my cell phone away from my body more often” did not change at the post-test. Although a lack of recognition of a need to change mobile phone behaviors may be viewed as challenging the effectiveness of the intervention, it is important to view each item as contributing to a broader construct.
Some additional limitations and future directions are also noteworthy. First, although the participants in this controlled study provided evidence that they changed their mobile phone use behaviors, and this post-intervention pattern appeared to be persistent (relative to the comparison group at one week), the endpoints measured were based on self-report. Future efforts to reduce mobile phone use could also incorporate more objective measures, such as phone bills or an application that records mobile phone use.61 Second, the participants were young adults and disproportionately female. As high-school and college-age students have the highest mobile phone usage,21 and some data indicates that females have more problematic mobile phone use patterns,50,53 we view the sample characteristics as a strength of this report. However, additional longitudinal studies will be necessary to further characterize the persistence of mobile phone attitudinal and behavior changes following interventions. Third, although those that did, and did not, complete the post-test were indistinguishable on fourteen measures, completers were both younger and more likely to view their mobile phone as a necessary expense. Further research will be necessary to determine if these group differences were Type I errors or if these characteristics are important indices for determining who will participate in longitudinal mobile phone investigations. Longitudinal research is inherently challenging and these findings should be viewed within this context, particularly the data obtained at the second post-test. Finally, this investigation was completed with a normative sample of young adults. More intense, and repeated, interventions may be needed for individuals with more excessive, and problematic, mobile phone attitudes and behaviors.27
In conclusion, this investigation provides evidence that mobile phone use can be modified, at least temporarily, in young adults.
BJ Piper: Formal analysis, software, writing—original draft, writing—reviewing & editing
SM Daily: Data curation, writing—original draft, writing—reviewing & editing
SL Martin: Writing—original draft, writing—reviewing & editing
MW Martin: Conceptualization, data curation, supervision, writing—original draft, writing—reviewing & editing
The efforts of the peer educators are gratefully acknowledged. BJ Piper is supported by the Health Resources Services Administration (D34HP31025). Software used for figures was provided by the National Institutes of Environmental Health Sciences (T32 ES007060- 31A1).
Disclosure of Interests
BJ Piper is a member of the editorial board of The Guthrie Journal, for which he receives no financial compensation. The authors declares that they have no additional competing interests.