University of Amsterdam Graduate School of Communication M.Sc. Entertainment Communication
Just take a ‘Moment’: Is this the cure
for smartphone addiction?
A study about the influence of apps that want to reduce smartphone
usage on the well-being of emerging adults
Name: Marieke Groenland
Student number: 12280763
E-Mail: marieke.groenland1@student.uva.nl
Thesis supervisor: dr. Susanne Baumgartner
Word Count: 10.561
Table of contents
Abstract ... 3
Introduction ... 4
Theoretical framework ... 7
Mental health apps and PSU ... 7
RSU-apps and time spend with smartphone ... 8
RSU apps and well-being ... 10
RSU-apps and FoMO ... 11
Gender and PSU ... 12
Motivation and RSU-apps ... 13
Method ... 14 Design ... 14 Sample ... 14 Procedure ... 15 Pre-test: Survey 1. ... 15 Post-test: Survey 2. ... 16 Measures ... 17 RSU-app. ... 17
Manipulation check and compliance... 18
PSU. ... 18
Well-being. ... 19
FoMO. ... 20
Purpose of smartphone use. ... 21
Time spent on smartphone. ... 22
Motivation. ... 22
Results ... 23
Randomization check ... 23
Manipulation check ... 23
Hypotheses 1: PSU ... 24
Hypotheses 2: Time spend ... 25
Hypotheses 3: Well-being ... 27
Hypothesis 5 (FoMO), 6 (gender) and 7 (motivation) ... 30
Discussion ... 32
Appendix: description exercises Phone Bootcamp ... 46
Abstract
This study investigated an app that wants to reduce smartphone usage (RSU-app) and the possible influences of this and similar apps on the time one spends on their smartphone, level of problematic smartphone use (PSU) and well-being. In addition, several factors that may influence compliance to RSU-apps have been taken into account: fear of missing out (FoMO), gender (female) and (intrinsic) motivation. This study tested the RSU-app ‘Moment’ and its course ‘Phone Bootcamp’. Several analyses indicated that compliance to the RSU-app assignments does not predict lower levels of PSU and also does not predict less time spent on the smartphone. The tests showed that participants did have significant lower levels of PSU after the two weeks of this study. However, no differences in PSU levels were found between the participants in the app and no-app condition. The same is true for the well-being of the participants: the analyses indicated a significant difference in well-being before and after the experiment, but no differences were found between the participants in the app and no-app condition. Moreover, the participants’ well-being did not increase, as the author would have expected, but in fact decreased after this study. Mediation analyses indicated that there was no mediating effect of time spent on smartphone and PSU on the relationship between compliance to the RSU-app and well-being. Finally, the predictors FoMO, gender (female) and intrinsic motivation turned out to be insignificant predictors of compliance to the RSU-app. As this study had some limitations, it is important to keep investigating RSU-apps in other circumstances. Future research can be done with other RSU-apps, another sample or with other predictors of well-being and compliance to these apps.
Introduction
Smartphones are playing an important role in people’s life. After the introduction of smartphones in 2007 (Twenge, Martin & Campbell, 2018), now (about a decade later) at least 90 percent of the population of every developed country owns a smartphone (Deloitte, 2017). Owning a smartphone has a lot of positive effects, such as facilitating social engagement and providing entertainment (Kim, Wang & Oh, 2016; Horwood & Anglim, 2019). However, using smartphones excessively may also be related to negative consequences, such as ‘phubbing’ (snubbing someone by checking one’s smartphone in the middle of a conversation; Balta, Emirtekin, Kircaburun & Griffiths, 2018), diminished daily work engagement (because of smartphone use the night before work; Lanaj, Johnson & Barnes, 2014), deteriorated physical conditions (such as obesity and diabetes; Griffiths, Lopez- Fernadez, Throuvala, Pontes & Kuss, 2018) and psychological conditions (e.g. anxiety and sleep disturbance; Van Velthoven, Powell & Powell, 2018).
This excessive smartphone use is already a widely discussed topic and is commonly named ‘problematic smartphone use’ (PSU). One very important factor influencing the severity of PSU is the time individuals spend on their smartphones (Rozgonjuk, Levine, Hall & Elhai, 2018). PSU is, according to Rozgonjuk et al. (2018), positively related to the screen time minutes of an individual. Taken everything into consideration, PSU can be defined as the inability to regulate one’s use of the smartphone (thus spending too much time on one’s smartphone), which eventually involves negative consequences in daily life: on physical, psychological, and social aspects (Billieux, 2012; Billieux, Mauarage, Lopez-Fernandez, Kuss & Griffiths, 2015; Van Deursen, Bolle, Hegner & Kommers, 2015).
The negative consequences of PSU can especially be linked to emerging adults (18 to 35-year olds), as this group is the largest group of smartphone owners and are the most dependent on their smartphone (28%), compared to older adults (24%; Pew Research Center,
2018). Younger target groups also use their smartphones a lot but are still dependent on their parents for using a smartphone and many children and adolescents cannot freely use their devices whenever they want (Anderson & Jiang, 2018).
Recently, several tools have been developed to counter PSU. The primary goals of these tools are to give users more insight into their smartphone use and try to prevent excessive use. Examples of these tools are, Apple’s ‘screen time’ tool, Instagram’s ‘your activity’ tool and applications that will help reduce your smartphone usage. These apps can be named ‘reduce smartphone usage apps’ (RSU-apps). Unfortunately, to the author’s knowledge, no studies exist to date that tested whether these apps actually help to reduce smartphone usage and increase well-being. This is in line with other studies that suspect that strategies promoting more ‘healthy’ smartphone behavior could have a positive impact on psychological well-being (Tangmunkongvorakul, et al., 2019), but research has not yet provided evidence to support this assumption (Donker et al., 2013; Plaza, Demarzo, Herrera-Mercadal, García-Campayo, 2013; Van Velthoven et al., 2018).
Since there are many new developments to promote healthy smartphone use, and there is a lack of evidence supporting these apps, the main aim of the current study is to investigate the effectiveness of a RSU-app on reducing the frequency of smartphone use, reducing problematic smartphone use, and its effect on well-being. Hence, the following main research question will be investigated:
RQ 1: To what extent does a RSU-app reduce time spent on the smartphone, and problematic smartphone use, and what is its influence on the well-being of emerging adults?
Furthermore, in order to study the effectiveness of a RSU-app, this study will also take into account several variables that might predict the compliance with the RSU-app. Typical RSU-
apps provide the user with several tasks and practices to reduce one’s smartphone use. Examples of these practices are setting goals about ‘screen time’ or number of pick-ups of the smartphone. Some RSU-apps also provide ‘courses’ to follow which consists of smartphone related challenges every day (e.g. ‘Lay your phone down for 30 minutes’).
Understanding factors that influence whether and how much someone complies with the tasks is important to understand the effectiveness of the app. First, an individual’s initial level of ‘fear of missing out’ (FoMO) will be taken into consideration. FoMO can be defined as the fear that significant others might be having rewarding experiences from which one is absent (Przybylski, Murayama, De Haan & Gladwell, 2013). It is important to investigate this variable, because FoMO is characterized by the desire to stay connected with what others are doing all the time and thus might make it more difficult to follow the instructions of a RSU-app successfully. Moreover, according to several studies, demographical factors like gender might influence the compliance with a RSU-app (Beranuy, Oberst, Carbonell & Chamarro, 2009; Van Deursen et al., 2015; Kim et al., 2016; Lopez-Fernandez et al., 2017; Carbonell, Chamarro, Oberst, Rodrigo & Prades, 2018; Horwood & Anglim, 2019; Tangmunkongvorakul et al., 2019; Oviedo-Trespalacios, Nandavar, Newton, Demant & Phillips, 2019). Finally, the level of motivation can play a role in an individual’s compliance with a RSU-app, as motivation is an important factor driving behavior change (Johnson, Deterding, Kuhn, Staneva, Stoyanov & Hides, 2016). Therefore, the following question will be investigated:
RQ 2: How does ‘fear of missing out’, gender and motivation to use the ‘reduce smartphone usage app’ predict compliance of emerging adults with using the ‘reduce smartphone usage app’?
In sum, this study will examine whether using a RSU-app can actually reduce people’s level of PSU and time they spend on their smartphones, and whether this in turn increases their well-being. Moreover, certain predictors of compliance with the RSU-app will be taken into account. As several studies indicated that the level of PSU is becoming higher globally (Carbonell et al., 2018; Oviedo-Trespalacios et al., 2019), investigating RSU-apps and the possible influencing variables can contribute to the health of people who are addicted to their smartphones.
Theoretical framework
Mental health apps and PSU
Mental health apps, such as RSU-apps, have implemented several mechanisms through which mental health support is being provided. First, the most common health behavior strategy of these apps is self-monitoring (Payne, Lister, West & Bernhardt, 2015). This means that mental health apps function as monitoring tools, as they enable users to set health related targets (Dute, Bemelmans & Breda, 2016). The app in turn sends reminders of ongoing goals and motivations (Bakker, Rickwood, Kazantis & Rickard, 2016). When users reach their goal, they receive rewards from the app (Bakker et al., 2016), which stimulates the adherence of daily practices (Plaza et al., 2013) and increases the motivation of users (Dute et al., 2016). Moreover, mental health apps provide psychoeducation, which for example consist of making people aware of monetary loss due to smoking or making them aware of the disadvantages of obesity (Rathbone & Prescott, 2017). In this way the awareness of their mental health problem can increase, which in turn facilitates their motivation to help reduce their mental health problems (Dute et al., 2016).
By offering several assignments, RSU-apps typically make use of both behavior change strategies: encouraging self-monitoring and creating awareness of their (problematic) smartphone use. RSU-apps make use of self-monitoring by enabling setting individual goals related to smartphone use (e.g. setting a maximum for picking up the smartphone). The RSU-app will in turn send push notifications to remind people of their goals or to motivate them to reach new goals. By doing this, the RSU-app creates awareness of their (problematic) smartphone use. The final goal of these strategies, used by RSU-apps, is thus to improve unhealthy smartphone use. Therefore, if an individual follows the RSU-app assignments, one can become more aware of their PSU and can self-monitor their problem better by following the assignments. Hence, the following hypothesis is formulated:
H1 A: Participants who use a RSU-app for two weeks will afterwards have lower PSU compared to participants who do not use a RSU-app for two weeks.
However, besides testing the difference between participants in the app and no-app condition, it is also important to look at the degree to which people in the experimental condition comply to the RSU-app, as their PSU levels might be influenced by the degree to which they actually follow the assignments. Therefore, the following hypothesis is formulated:
H1 B: Participants who comply more to the RSU-app assignments for two weeks will afterwards have lower PSU.
RSU-apps and time spend with smartphone
By making use of the abovementioned behavior change strategies, the main aim of RSU-apps is to reduce the time people spend on their smartphone. As the amount of time spend on one’s
smartphone is part of one’s PSU level, several studies indicate that the greater time one spends on their smartphone can lead to PSU (Haug, Castro, Kwon, Filler, Kowatsch & Schaud, 2015; Lopez-Fernandez et al., 2017). In addition, tracking screen time minutes is positively associated with PSU (Rozgonjuk et al., 2018). Hence, in this study only the minutes spend on smartphones will be taken into account. As RSU-apps aim to reduce the time people spend on their smartphone (and thus try to reduce PSU), following the RSU-app assignments can have a positive influence on the time one spends on their smartphone. Therefore, it is expected that individuals who conform more to the assignments, will spend less time on their smartphone.
H2 A: Participants who use a RSU-app for two weeks will spend less time (per week, in minutes) on their smartphone compared to participants who do not use a RSU-app for two weeks.
However, besides testing the difference between participants in the app and no-app condition, it is also important to look at the degree to which people in the experimental condition comply to the RSU-app, as the time they spend on their smartphone might be influenced by the degree to which they actually follow the assignments. Therefore, the following hypothesis is formulated:
H2 B: Participants who comply more to the RSU-app assignments for two weeks, will afterwards spend less time (per week, in minutes) on their smartphone.
RSU apps and well-being
According to two studies there is a link between the time spend on digital media (computers, smartphones, tablets etc.) and a person’s well-being (Twenge et al., 2018; Twenge & Campbell, 2019). Namely, adolescents who spent more time on their smartphones and other electronic communication devices (e.g. social media and gaming) and less time on non-screen activities (e.g. in-person interaction and sports) had a lower psychological well-being (Twenge et al., 2018). As the assignments of RSU-apps stimulate to spend less time on smartphones, and more time on non-screen activities, it is expected that participants who comply more to the assignments will afterwards have higher well-being.
H3 A: Participants who use a RSU-app for two weeks will afterwards have higher well-being compared to participants who do not use a RSU-app for two weeks.
However, besides testing the difference between participants in the app and no-app condition, it is also important to look at the degree to which people in the experimental condition comply to the RSU-app, as their well-being might be influenced by the degree to which they actually follow the assignments. Therefore, the following hypothesis is formulated:
H3 B: Participants who comply more to the RSU-app assignments for two weeks will afterwards have higher well-being.
In addition, individuals who are light digital media users (less than one hour a day) reported substantially higher well-being than heavy digital media users (more than five hours a day; Twenge & Campbell, 2019). As mentioned above, as RSU-apps with its assignments stimulate to spend less time on smartphones, this can have a positive effect on an individual’s
well-being. Therefore, it is expected that the positive effect of the RSU-app on well-being is mediated by the reduced time one spends on their smartphone.
H4 A: The relationship between compliance to the RSU app and well-being is mediated by time spent on smartphone: completing more assignments of the app leads to less time spend on the smartphone and this in turn leads to increases in well-being.
Moreover, one longitudinal study showed that the more problematic an individual’s smartphone use was, the less social support they got from their environment, which in turn resulted in an increase of smartphone addiction (Herrero, Urueña, Torres & Hildalgo, 2019). This increase in problematic use of smartphones can, in turn, be linked to a lower well-being, (Horwood & Anglim, 2019; Tangmunkongvorakul, et al., 2019; Volkmer & Lermer, 2019). Hence, reducing one’s level of PSU might lead to increases in one’s level of well-being. Therefore, it is expected that the positive effect of the RSU-app on well-being is also mediated by reduced levels of PSU.
H4 B: The relationship between compliance to the RSU app and well-being is mediated by PSU: completing more assignments of the app leads to less PSU and this in turn leads to increases in well-being.
RSU-apps and FoMO
Since it is expected that RSU-apps have positive effects on one’s well-being, it is important to understand factors that make it more likely that people actually use the app and comply to its instructions. One of the factors that might have an influence on the compliance to RSU-apps is FoMO. As discussed before, FoMO can be described as the fear of missing out on
important events or experiences (Przybylski et al., 2013). Missing out on these important events, can thus trigger a negative feeling, as several studies indicate that FoMO can have a negative effect on one’s well-being (Przybylski et al., 2013; Stead & Bibby, 2017; Griffiths et al., 2018).
This might lead to a higher level of smartphone use and, in some cases, eventually PSU (Chotpitayasunondh & Douglas, 2016; Elhai, Levine, Dvorak & Hall, 2016; Van Velthoven et al., 2018; Griffiths et al., 2018; Wolniewicz, Tiamiyua, Weeks & Elhai, 2018). Therefore, FoMO might have an influence on the effectiveness of a RSU-app, as this trait makes it hard for individuals to put down their smartphone. The RSU-app encourages people to use their smartphone less often (e.g. both in picking up the phone less and spending less time on the smartphone), which can trigger their FoMO. Therefore, it can be suggested that people with higher levels of FoMO may find it harder to conform to the assignments of RSU-apps. Hence the following hypothesis is formulated:
H5: Participants with higher levels of FoMO will comply less to the RSU-app assignments.
Gender and PSU
Besides FoMO, gender may have an influence on whether individuals will comply to the RSU-app assignments. According to several studies, females use their smartphone to a greater extent (than men) for social purposes, such as maintaining social relationships by using social networks, internet services and talking on the phone (Lee, Chang, Lin & Cheng, 2014; Van Deursen et al., 2015; Carbonell et al., 2018; Volkmer & Lermer, 2019). Men, on the other hand, are more task-oriented users and use their smartphones more for completing tasks by talking on the phone or use their phones for gaming (Lee et al., 2014; Carbonell et al., 2018; Volkmer & Lermer, 2019). Individuals who use their smartphones more for social purposes
develop smartphone habits faster, because of the notifications on their smartphone (Van Deursen et al., 2015).
This is in line with several other studies that claim that females experience higher PSU levels compared to males (Beranuy, Oberst, Carbonell & Chamarro, 2009; Van Deursen et al., 2015; Kim et al., 2016; Lopez-Fernandez et al., 2017; Carbonell et al., 2018; Horwood & Anglim, 2019; Tangmunkongvorakul et al., 2019; Oviedo-Trespalacios et al., 2019). As females in general use their smartphones more for social purposes, it can be assumed that they developed greater smartphone habits, and this might make it harder for them to comply to the RSU-app assignments. Hence, the following hypothesis is formulated:
H6: Female participants will comply less with the RSU-app assignments, compared to male participants.
Motivation and RSU-apps
Finally, the level of motivation can play a role in whether an individual will conform to the RSU-app assignments. It is important to take this variable into account as health and well-being strongly depend on an individual's healthy behavior and motivation can be seen as a main factor driving behavior change (Johnson et al., 2016). According to Ryan and Deci (2000) motivation can be distinguished in two different types. First, individuals can be intrinsically motivated to do something. This means that they perform a kind of behavior because they think it is inherently interesting or enjoyable (Ryan & Deci, 2000). Second, individuals can be extrinsically motivated to perform a behavior, which means that they perform a behavior for some separable consequence (such as pressure or rewards; Ryan & Deci, 2000). However, intrinsically motivated individuals are more successful in performing preventive health behavior (Sundar, Bellur & Jia, 2012) as this type of motivation is a key
factor for assignment adherence (Yoganathan & Kajanan, 2014). Therefore, the following hypothesis is formulated:
H7: Participants who are intrinsically motivated to use a RSU-app, will comply more to the RSU-app assignments.
Figure 1 provides an overview of hypotheses.
Figure 1. Overview of hypotheses 1B, 2B, 3B, 5, 6, 7.
Method
Design
This study had a two-group repeated measures experimental design, with one experimental condition (RSU-app) and one control condition (no RSU-app).
Sample
To gather participants a convenience sample in combination with a snowball sample was used. The survey was distributed by posting it on personal social media pages and friends and family were asked to spread the survey to emerging adults they know. In total 145 emerging
adults (18 to 35-year olds) were gathered. However, of these 145 participants, only 78 filled out a real email-address, which was necessary to send the second questionnaire. After that only 70 participants filled out both questionnaires. Therefore, the sample of this study consists of 70 participants (16 males and 54 females), aged (M = 23.06, SD = 3.28, min = 18, max = 35). In addition, the greatest part of these participants (47 of 70) had a high education level (i.e. bachelor or master’s degree). Of this sample, 57 participants were Dutch, and the others originated from other countries (such as China, Afghanistan and Germany). Mostly Dutch participants were gathered as this study took place in the Netherlands and this country has the highest smartphone penetration worldwide (namely 93% of Dutch people owns a smartphone; Consultancy.nl, 2018). In addition, in 2018, already 30% of the Dutch youngsters called themselves addicted to their smartphone (Van Mersbergen, 2018). Therefore, PSU can be seen as a problem among Dutch emerging adults.
Procedure
Before conducting the main study, a small pilot study was conducted to test if all the questions were understandable and if Qualtrics was randomizing the conditions correctly. Subjects were randomly assigned to the app and no-app conditions in the first survey. The participants could win a Bol.com voucher or Fashioncheque (to the value of 20 euros) by participating to this study. In addition, all participants were rewarded with a free code to use the ‘Moment’ app for free.
Pre-test: Survey 1. At the beginning of the experiment, participants were asked to fill out an informed consent, in which they were made aware that they took part in this study voluntarily and anonymously. After this the participants were randomly assigned to either the experiment condition or the control condition. The participants in the experimental condition
received a slightly different informed consent: in this version they were instructed that they had to download the app ‘Moment’. After this, they received an explanation of the experimental condition, in which the participants were instructed to download the app and unlock the ‘Moment coach’. The participants had to confirm they downloaded the app, before they could continue with answering the other questions. Furthermore, the participants in the experimental condition were asked to use the app and follow the course ‘Phone Bootcamp’ for two weeks.
After that all participants (both experimental and control condition) were asked to provide their demographics (gender, age, level of education and country of origin). Both conditions were asked to fill out questions about the average time (per week, in minutes) they had spent on their smartphone, PSU, FoMO and well-being. The experimental condition was, in addition, asked to fill out their motivation to use the ‘Moment’ app.
Post-test: Survey 2. The second survey was sent two weeks after the participants completed the first questionnaire. This survey was almost the same for all the participants (both experimental and control condition). The participants were asked to provide the average time they spent on their smartphone again. In addition, the participants were asked to fill out the PSU and well-being questionnaire. FoMO was only measured during the pre-test because FoMO can be seen as a trait: this will not change over a timespan of two weeks. Finally, the participants were asked to answer the manipulation check about whether they used the app ‘Moment’ or not, (which would indicate whether they were in the experimental condition) and were asked some general questions about the app.
Measures
RSU-app. In this study, compliance to the RSU-app was measured by the absence (control condition) or presence (experimental condition) of instructing the participants to use the RSU-app ‘Moment’ for two weeks. Participants in the control condition were only asked to complete the pre- and post-test, with a time of two weeks in between the surveys (because this is the same time span of the ‘Moment Phone Bootcamp’).
Moment. This RSU-app is chosen for this study, because this app is available for both
iOS and Android, does not ask the users to give any personal information, is very easy to use and was worldwide one of the most downloaded and positively rated RSU-apps. Namely, out of 5 stars ‘Moment’ got 4.4 (based on 1,8K ratings; App Store, n.d.). In addition, this app is comparable to other RSU-apps like ‘Flipd’ (4.4 stars, 47 ratings) and ‘Space’ (4.2 stars, 81 ratings: App Store, n.d.), as they provide similar exercises like tracking smartphone use, setting screen time limits and providing overviews.
However, ‘Moment’, is the only app which provides a ‘coach’ which helps users to follow different courses that will help the user to reduce their smartphone use. Examples of these courses are ‘Attention span’ (16 days), ‘Better sleep’ (14 days), ‘Mini detox’ (4 days) and the ‘Phone Bootcamp’ (14 days). For this study the last one was tested, because this course is specifically focused on reducing smartphone usage. See Appendix for a list of descriptions of the ‘Phone Bootcamp’ exercises. Moreover, the ‘Phone Bootcamp’ has a longer timespan compared to the ‘Mini detox’ (which has the same aim as the ‘Phone Bootcamp’, namely reducing smartphone usage). The ‘Mini detox’ can be seen as too radically for beginners of RSU-apps, because it asks the users to put their phone away for two whole days. Additionally, it must be noted that the Moment coach is a subscription and
normally costs money, but for this study Moment made an exception and provided a free code in exchange for providing an overview of the main findings of this study.
Manipulation check and compliance. The manipulation check was done at the beginning of the second questionnaire by asking the following question to the participants: “To what extent did you use the app 'Moment' to help reduce your smartphone usage?”. The participants could indicate on a five-point Likert scale, ranging from 0 (‘never’) to 5 (‘daily’). If ‘never’ was selected, the survey would skip all the questions concerning the app. However, if one of the other options were selected, the participants were provided a list of the ‘Phone Bootcamp’ exercises and were asked to check the boxes of the exercises they performed. In this way, the total sum of the boxes checked formed together the score of compliance to the RSU-app assignments (in the result section named ‘total Phone Bootcamp’). The participants could get a minimum score of 0 and maximum score of 14, with higher scores indicating greater compliance to the RSU-app assignments (M = 4.77, SD = 4.04). In addition, also some general questions about the app were asked such as “Did you like the ‘Moment’ app?” and “How useful do think this app was?”.
PSU. The variable PSU was measured by using the smartphone addiction scale compiled by Kwon, Kim, Cho and Yang (2013). This questionnaire was chosen, because the reliability is excellent (α = 0.91). The PSU-scale consist of 10 items on a six-point Likert scale ranging from ‘strongly disagree’ (scored as 1) to ‘strongly agree’ (scored as 6). Examples of questions are: “I won’t be able to stand not having a smartphone” and “The people around me tell me that I use my smartphone too much”. However, question 8 was slightly changed for this study by adding a somewhat new social medium which is very often used nowadays: Instagram. This question was changed into: “I was constantly checking my
smartphone so as not to miss conversations or pictures of other people on Twitter, Facebook or Instagram”.
The participants were asked to indicate how often the ten statements would apply to them in the past week. The sum of the item scores is the overall measure of PSU (ranging from 10 to 60), with higher scores indicating a higher level of PSU. This applies to both of the surveys: the one the participants had to fill out in the pre-test (from which the variable is named ‘PSU 1’) and the one they had to fill out in the post-test (from which the variable is named ‘PSU 2’). An exploratory factor analysis with Oblimin rotation indicated that the ‘PSU 1’ scale was unidimensional, explaining 61.54% of the variance. The 10-item scale also proved reliable in this study as indicated by a Cronbach’s Alpha of .82 (M = 31.63, SD = 10.28). Another exploratory factor analysis with Oblimin rotation indicated that the ‘PSU 2’ scale was also unidimensional, explaining 62.17% of the variance. The 10-item scale also proved reliable as indicated by a Cronbach’s Alpha of .80 (M = 29.89, SD = 9.37). Moreover, a variable was constructed to calculate the score differences between the post- and pre-test: ‘total PSU’ (M = -1.74, SD = 7.86).
Well-being. The level of well-being was measured by using the World Health Organization well-being index, consisting of 5 items on a six-point Likert scale ranging from ‘at no time’ (scored as 0) to ‘all of the time’ (scored as 5; Topp, Østergaard, Søndergaard & Bech, 2015). The participants were asked to indicate how often the five statements could be applied to them in the past week. Examples of questions are: “I have felt cheerful and in good spirits” and “My daily life has been full of things that were interesting to me”. This questionnaire was chosen because it is able to show overall change along the continuum of well-being and its applicability across study fields is very high (Topp et al., 2015). In
addition, the reliability of the questionnaire is good (α = 0.87; Saipanish, Lotrakul, Sumrithe, 2009).
The score of this questionnaire was calculated in the following way: the raw score ranges from 0 to 25), with lower scores indicating a lower level of well-being. This applies to both of the surveys: the one the participants had to fill out in the pre-test (from which the variable is named ‘well-being 1’) and the one they had to fill out in the post-test (from which the variable is named ‘well-being 2’). An exploratory factor analysis with Oblimin rotation indicated that the ‘well-being 1’ scale was unidimensional, explaining 59.44% of the variance. The 5-item scale also proved reliable in this study as indicated by a Cronbach’s Alpha of .83 (M = 16.09, SD = 4.38). Another exploratory factor analysis with Oblimin rotation indicated that the ‘well-being 2’ scale was also unidimensional, explaining 55.40% of the variance. The 5-item scale also proved reliable as indicated by a Cronbach’s Alpha of .79 (M = 13.97, SD = 3.53). Moreover, a variable was constructed to calculate the score differences between the post- and pre-test: ‘total well-being’ (M = -2.11, SD = 4.22).
FoMO. The variable FoMO was measured by using the questionnaire by Wegmann, Oberst, Stodt and Brand (2017). This questionnaire was chosen for this study because it measures both ‘offline’ experiences (related to the fear of missing out on ‘offline’ meetings with friends) as ‘online’ experiences (related to the fear of missing out on anything ‘online’, e. g. social media). Moreover, the reliability of both subscales was good (offline FoMO: α = 0.82; online FoMO: α = 0.81; Wegmann et al., 2017). The FoMO scale consists of 12 items on a seven-point Likert scale from ‘totally disagree’ (scored as 1) to ‘totally agree’ (scored as 5). Examples of the items are “When I miss out on a planned get-together it bothers me” (offline FoMO) and “I continuously consult my smartphone, in order not to miss out on
anything” (online FoMO). Moreover, the offline FoMO subscale consists of 5 items and the online FoMO subscale consists of 7 items.
An exploratory factor analysis with Oblimin rotation indicated that the scale was not unidimensional, but instead was loading on 3 factors, explaining together 68.37% of the variance. However, the pattern matrix showed that the third factor consisted of two negative items (“I fear my friends have more rewarding experiences than me” and “I fear others have more rewarding experiences than me”). These negative items belonged to the ‘offline FoMO’ scale and were subtracted from the other items. Therefore, that subscale consists now of 3 items. After excluding the negative items, a new factor analysis with Oblimin rotation indicated that there were two unidimensional scales (offline and online), with offline FoMO explaining 14.38% of the variance and online FoMO explaining 47.52% of the variance (together explaining 61.9% of the variance). Both sub-scales were proved reliable in this study by a Cronbach’s Alpha of .80 (offline FoMO; M = 10.4, SD = 4.14) and .86 (online FoMO; M = 19.73, SD = 8.32). Moreover, the sum of the item scores is the overall measure of offline FoMO (ranging from 3 to 21) and online FoMO (ranging from 7 to 49), with higher scores indicating higher levels of offline and online FoMO.
Purpose of smartphone use. This variable was measured by two items. The participants were asked to indicate on a five-point Likert scale ranging from ‘not at all’ (0) to ‘very much’ (5), to what extent they use their smartphone for social purposes (e.g. social networking, calling and maintaining social relationships). In addition, the participants were asked on the same scale to indicate to what extent they use their smartphone for passing time and relaxation purposes (e.g. entertainment such as games or movies). The participants scored an average score of 2.76 (SD = .91) for social purposes and an average score of 2.36 for entertainment purposes (SD = 1.13).
Time spent on smartphone. The variable ‘time spent on smartphone’ was measured using one item. Participants were asked to indicate what their average smartphone usage was in the past week. Participants who had an iPhone were asked to check this by going to their ‘screen time’ overview in their phone settings. Participants who did not own an iPhone were asked to make an educated guess. Moreover, the participants had to indicate whether they checked their specific screen time or that they made a guess. In the pre-test, 50% of the participants looked up their specific screen time and the other 50% made an educated guess. In the post-test, 61.4% of the participants looked up their specific screen time and 38.6% of the participants made an educated guess. The participants scored an average of 1519.53 minutes per week (SD = 811.27) in the pre-test and an average of 1459.39 minutes per week in the post-test (SD = 1119.83). Moreover, a variable was constructed to calculate the score differences between the post- and pre-test: ‘total time spent’ (M = -60.14, SD = 1208.49).
Motivation. Finally, to check whether the participants (of the experimental condition) were intrinsically or extrinsically motivated to start using the app ‘Moment’, two questions were asked. First, to check whether participants were intrinsically motivated to use the app, they were asked to indicate on a five-point Likert scale, ranging from ‘not motivated at all’ (0) to ‘extremely motivated’ (5), how motivated they were to use the app ‘Moment’ to help reducing their smartphone usage for themselves. Second, to check whether participants were extrinsically motivated to use the app, they were asked to indicate on the same scale, how motivated they were to use the app ‘Moment’ for the rewards they may win by participating to this study. The participants scored an average of 2.77 on level of intrinsic motivation (SD = 1.07) and an average of 1.73 on level of extrinsic motivation (SD = .92).
Results
Analyses were conducted using a final sample of 70 participants in total.
Randomization check
In order to check if participants’ age was comparable over the two conditions, a one-way ANOVA was conducted. This ANOVA had ‘condition’ (app vs. no-app) as independent variable, and age as dependent variable. The ANOVA showed that participants’ mean age in the app condition (M = 22.85, SD = 2.95) was not significantly different from participants’ mean age in the no-app condition (M = 23.18, SD = 3.50), F (1, 68) = .17, p = .68. Moreover, to check whether the participants’ gender was comparable over the two conditions (app-condition: 8 males, 18 females and the no app-(app-condition: 8 males, 36 females), a Pearson’s chi-square test of contingencies with (α = .05) was conducted. This chi-square test was not significant, 2(1, N = 70) = 1.47, p = .25, which means that the gender distribution was comparable over the conditions.
Manipulation check
To check whether the participants perceived the conditions as intended, an independent samples t-test was conducted with ‘condition’ (app vs. no app) as independent variable, and ‘manipulation check’ as dependent variable, containing the following question: “To what extent did you use the app 'Moment' to help you reduce your smartphone usage?”. A significant effect of ‘manipulation check’ was found on ‘condition’, t(-3.85) = 40.50, p < .001. The participants in the experimental condition used the RSU-app significantly more (M = 1.46, SD = 1.27) than the participants in the control condition (M = .36, SD = .92). In addition, it was checked if only participants in the app condition downloaded the app and this
was confirmed. The t-test and the abovenamed check show that the manipulation of the RSU-app was successful.
Hypotheses 1: PSU
To test hypothesis 1A, which suggests that the participants who used a RSU-app for two weeks, would have lower levels of PSU compared to participants who do not use such an app for two weeks, a one-way repeated measures analysis of variance (ANOVA) was conducted. PSU at t1 (‘PSU 1’) and t2 (‘PSU 2’) were used as the within-subject factor, and condition (app vs. no-app) as the between subject factor. The ANOVA showed that there were significant differences in ‘PSU 1’ (M = 31.63, SD = 10.28) and ‘PSU 2’ (M = 29.89, SD = 9.37), F (1, 68) = 4.35, p = .04, partial 2 = .06. This indicates that participants had significant lower levels of PSU after the two weeks of study participation. However, the ANOVA indicated that there are no significant differences in PSU levels between the participants in the app condition (‘PSU 1’: M = 32.15, SD = 10.53, ‘PSU 2’: M = 29.04, SD = 9.65) and the participants in the no-app condition (‘PSU 1’: M = 31.32, SD = 10.24, ‘PSU 2’: M = 30.39, SD = 9.27), F (1, 68) = 1.27, p = .26. Thus, participants who used a RSU-app for two weeks did not have lower levels of PSU, compared to participants who did not use such an app. Hence, hypothesis 1A is rejected.
Furthermore, to test hypothesis 1B, which suggests that participants who would comply more to the RSU-app assignments would afterwards have lower levels of PSU, a standard simple regression analysis was conducted. To test this hypothesis, only the participants in the experimental condition (N = 26) were used, as this group was only instructed to use the RSU-app. For this analysis, ‘total PSU’ (which indicates the score difference between the participants’ PSU level before and after the two weeks of this study) was used as dependent variable for this analysis and ‘total Phone Bootcamp’ (which indicates
the total of the completed assignments of the Phone Bootcamp) was used as independent variable. The analysis pointed out that compliance to RSU-app assignments accounted for a non-significant .80% of the variability of PSU, R2 = .008, F (1.24) = .20, p = .86, B = .22, SE = .49, = .09, t = .45, CI [-.80, 1.24]. Thus, compliance to the RSU-app assignments does not predict lower levels of PSU after the two weeks, which means that hypothesis 1B is rejected.
Hypotheses 2: Time spend
To test hypothesis 2A, which suggests that the participants who use a RSU-app for two weeks, would spend less time (per week, in minutes) on their smartphone compared to participants who do not use such an app for two weeks, a one-way repeated measures analysis of variance (ANOVA) was conducted. Time spent at t1 (‘time spent 1’) and t2 (‘time spent 2’) were used as the within-subject factor, and condition (app vs. no-app) as the between subject factor. Moreover, ‘check time spent 1’ and ‘check time spent 2’ were used as covariates.
The ANOVA showed that there were no significant differences in ‘time spent 1’ (M = 1519.51, SD = 811.27) and ‘time spent 2’ (M = 1459.39, SD = 1119.83), F (1, 66) = 1.47, p = .23. Moreover, there were no significant differences in the time participants had spent on their smartphone, between the participants in the app condition (‘time spent 1’: M = 1316.81, SD = 527.32, ‘time spent 2’: M = 1241.35, SD = 521.55) and the participants in the no-app condition (‘time spent 1’: M = 1639.32, SD = 924.62, ‘time spent 2’: M = 1588.23, SD = 1344.76), F (1, 66) = .21, p = .65. This is also the case when the covariates were taken into account: ‘check time spent 1’, F (1, 66) = .86, p = .36, and ‘check time spent 2’, F (1, 66) = 0.0, p = .96. Thus, participants who used a RSU-app for two weeks did not spend less time on their smartphone after the two weeks, compared to participants who did not use such an app. Hypothesis 2A is, therefore, rejected.
Moreover, to test hypothesis 2B, which suggests that participants who would comply more to the RSU-app assignments would spend less time (per week, in minutes) on their smartphone afterwards, a hierarchical multiple regression analysis was employed. To test this hypothesis, only the participants in the experimental condition (N = 26) were used, as this group was only instructed to use the RSU-app. For this analysis ‘total time spent’ (which indicates the score difference between the time participants spent on their smartphone before and after the two weeks of this study) was used as dependent variable and ‘total Phone Bootcamp’ was used as the independent variable. The variables ‘check time spent 1’ and ‘check time spent 2’, which indicate whether the participants specifically checked the time spent on their smartphone or that they made an educated guess, were used as covariates in the analysis. On step one of the hierarchical regression analysis, ‘check time spent 1’ and ‘check time spent 2’ accounted for a non-significant 5% of the variance in ‘total time spent’, R2 = .05, F (2, 23) = .61, p = .55. On the second step, ‘total Phone Bootcamp’ was added to the analysis and accounted for an insignificant additional 7.3% of the variance in ‘total time spent’, R2 = .073, F (1, 22) = 1.82, p = .19. In combination the three predictor variables explained 12.3% of the variance in ‘total time spent’. Thus, compliance to the RSU-app assignments does not predict less time spend on smartphone after the two weeks. Hence, hypothesis 2A is rejected. Unstandardized regression coefficients (B), coefficients standard error (SE B), standardized regression coefficients (), and t- and p-values of the predictors in the regression model are reported in Table 1.
Table 1.
Unstandardized regression coefficients (B), coefficients standard error (SE B), standardized regression coefficients (), and t- and p-values of each predictor in a regression model predicting ‘total time spend’
Variable B [95% CI] SE B t p Step 1
Check time spent 1 Check time spent 2 Step 2
Check time spent 1 Check time spent 2 Total phone bootcamp
-45.83 [-580.48, 488.81] -239.00 [-922.92, 444.92] -71.09 [-599.35, 456.98] -152.80 [839.35, 533.78] -31.35 [-79.49, 16.80] 258.45 330.61 254.63 331.06 23.21 -.05 -.19 -.07 -.12 -.27 -.18 -.72 -.28 -.46 -.35 .86 .48 .78 .65 .19 Note. N= 26, CI = confidence interval
*p < .05
Hypotheses 3: Well-being
To test hypothesis 3A, which suggests that the participants who use a RSU-app for two weeks, would have higher well-being compared to participants who do not use such an app for two weeks, a one-way repeated measures analysis of variance (ANOVA) was conducted. Well-being at t1 (‘well-being 1’) and t2 (‘well-being 2’) were used as the within-subject factor, and condition (app vs. no-app) as the between subject factor. The ANOVA shows that there are significant differences in ‘well-being 1’ (M = 16.09, SD = 4.38) and ‘well-being 2’ (M = 13.97, SD = 3.53), F (1, 68) = 20.98, p < .001, partial 2 = .24. However, the ANOVA indicated that there were no significant differences in well-being levels between the participants in the app condition (‘well-being 1’: M = 16.08, SD = 4.59, ‘well-being 2’: M = 12.81, SD = 2.48) and the participants in the no-app condition (‘well-being 1’: M = 16.09, SD = 4.30, ‘well-being 2’: M = 14.66, SD = 3.89), F (1, 68) = 3.20, p = .08. Thus, participants who used a RSU-app for two weeks did not have higher levels of well-being compared to participants who did not used such an app. Hence, hypothesis 3A is rejected.
Furthermore, to test hypothesis 3B, which suggests that participants who would comply more to the RSU-app assignments would afterwards have higher well-being, a standard simple regression analysis was conducted. To test this hypothesis, only the
participants in the experimental condition (N = 26) were used, as this group was only instructed to use the RSU-app. For this analysis ‘total well-being’ (which indicates the score difference between the participants’ well-being level before and after the two weeks of this study) was used as dependent variable for this analysis and ‘total Phone Bootcamp’ was used as the independent variable. The regression pointed out that compliance to RSU-app assignments accounted for a non-significant 5.6% of the variability of well-being, R2 = .06, F (1, 24) = 1.43, p = .24, B = -.29, SE = .24, = -.24, t = -1.2, CI [-.79, .21]. This indicates that, compliance to the RSU-app assignments does not predict higher well-being after the two weeks. Hence, hypothesis 3B is also rejected.
Hypotheses 4: Mediations
To test hypothesis 4A, which suggests that the relationship between compliance of the RSU-app and well-being is mediated by ‘total time spent’ (completing more assignments of the RSU-app would lead to less time spend on smartphone and this in turn leads to increases in well-being), a mediation analysis was done, by using the PROCESS procedure created by Hayes (2013). The standardized regression coefficient of the relationship between ‘total Phone Bootcamp’ and ‘total time spent’ was not significant, B = -11.18, SE = 42.70, t(1, 68) = -.26, p = .79, as Figure 2 shows. However, the direct effect of ‘total Phone Bootcamp’ and ‘total well-being’ was significant, but this was the opposite direction as expected, B = .39, SE = .14, t(1, 68)= -2.72, p = .008. Moreover, this relation stayed significant when ‘total time spent’ was added, B = -.38, SE = .14, t(1, 68)= -2.69, p = .009. The relationship between ‘total time spent’ and ‘total wellbeing’ was not significant, B = .0003, SE = .0004, t(2, 67)= .80, p = .43. The standardized coefficient for the indirect effect was B = -.004 but was not significant because the confidence interval passed zero, CI [-.03, .03]. This points out that there is no mediating
effect of ‘total time spent’ on the relationship between ‘total Phone Bootcamp’ and ‘total well-being’. Therefore, hypothesis 4A is rejected.
Figure 2. Mediation model of ‘time spend’ on the relation between ‘compliance with RSU-app’ and ‘well-being’.
Note. ** p < .001
Moreover, to test hypothesis 4B, which suggests that the relationship between compliance of the RSU-app and well-being is mediated by ‘total PSU’ (completing more assignments of the app would lead to less PSU and this in turn leads to increases in well-being), a mediation analysis was done by using the PROCESS procedure created by Hayes (2013). The standardized regression coefficient of the relationship between ‘total Phone Bootcamp’ and ‘total PSU’ was not significant, B = -.16, SE = .28, t(1, 68) = .-.60, p = .56, as Figure 3 shows. However, the direct effect of ‘total Phone Bootcamp’ and ‘total well-being’ was significant, but in the opposite direction as expected as mentioned before, B = -.39, SE = .14, t(1, 68)= -2.72, p = .008. Moreover, this relation also stayed significant when ‘total PSU’ was added, B = -.37, SE = .06, t(2, 67)= -1.33, p = .01. The relationship between ‘total PSU’ and ‘total wellbeing’ was not significant, B = .08, SE = .56, t(2, 130)= 1.56, p = .19. The standardized coefficient for the indirect effect was B = -.01 but was not significant because the confidence interval passed zero, CI [-.09, 0.05]. This points out that there is no mediating
effect of ‘total PSU’ on the relationship between ‘total Phone Bootcamp’ and ‘total well-being’. Therefore, hypothesis 4B is also rejected.
Figure 3. Mediation model of ‘PSU’ on the relation between ‘compliance with RSU-app’ and ‘well-being’.
Note. ** p < .001
Hypothesis 5 (FoMO), 6 (gender) and 7 (motivation)
To test the last three hypotheses, which state that participants with higher levels of FoMO will comply less to the RSU-app assignments (H5), that female participants will comply less to the RSU-app assignments (compared to male participants) because females use their smartphones more for social purposes (H6) and that participants who are intrinsically motivated to use a RSU-app would comply more to the RSU-app assignments (H7), a standard multiple regression analysis was done. To test these hypotheses, only the participants in the experimental condition (N = 26) were used, as this group was only instructed to use the RSU-app.
For this regression analysis, ‘offline FoMO, ‘online FoMO, ‘social purposes’ (which indicates to what extent the participants used their smartphones for social purposes), ‘gender’, (a new computed) interaction variable ‘gender*social-purposes’, ‘intrinsic motivation’ and ‘extrinsic motivation’ were used as independent variables. In addition, ‘total Phone
Bootcamp’ was used as the dependent variable. Together all the independent variables accounted for a non-significant 33% of the variability in ‘total Phone Bootcamp’, R2 = .33 F (7.18) = 1.25, p = .33.
Furthermore, ‘offline FoMO’ accounted for a non-significant .24 % of the variability in ‘total Phone Bootcamp’, B = -.06, SE = .27, = -.05 , t(7, 18) = -.21, p = .84, CI [-.61, .50], and ‘online FoMO accounted for a non-significant .50% of the variability in ‘total Phone Bootcamp’, B = -.04, SE = .12, = -.08 , t(7, 18) = -.3, p = .77, CI [-.30, .22]. Thus, participants with higher levels of ‘offline FoMO’ and ‘online FoMO’ do not comply less to the RSU-app assignment, and hypothesis 5 is rejected.
Moreover, ‘gender’ accounted for a non-significant 15.37% of the variability in ‘total Phone Bootcamp’, B = 8.95, SE = 4.95, = 1.04, t(7, 18) = 1.81, p = .09, CI [-.45, 19.34], ‘social purposes’ accounted for a non-significant 2.62% of the variability in ‘total Phone Bootcamp’, B = 2.64, SE = 3.79, = .64 t(7, 18) = .70, p = .50, CI [-5.32, 10.60], and ‘gender*social-purposes’ accounted for a non-significant 5.66% of the variability in ‘total Phone Bootcamp’, B = -2.17, SE = 2.09, = -1.27 t(7, 18) = -1.04, p = .31, CI [-6.55, 2.22]. This indicates that both ‘gender’ and ‘social purposes’ were no predictors for compliance to the RSU-assignments, and hypothesis 6 is rejected.
Finally, ‘intrinsic motivation’ accounted for a non-significant .05% of the variability in ‘total Phone Bootcamp’, B = -.07, SE = .76, = -.02 t(7, 18) = -.09, p = .93, CI [-1.67, 1.53], and ‘extrinsic motivation’ accounted for a non-significant 5.20% of the variability in ‘total Phone Bootcamp’, B = -1.01, SE = 1.01, = -.23 t(7, 18) = -.10, p = .33, CI [-3.13, 1.12]. Thus, participants who were more intrinsically or extrinsically motivated to use the RSU-app did not comply more to the RSU-app assignments, and hypothesis 7 is also rejected.
Discussion
The main focus of this study was to investigate a relative new development in the area of smartphones: RSU-apps. To date, to the author’s knowledge, no studies have been done to investigate whether these apps actually work. Therefore, this study looked at the main goals of RSU-apps and whether these goals can be reached: reducing the time one spends on their smartphone, reducing PSU, and increasing one’s well-being (RQ 1). In order to study this relation, this study also took into account several important variables that might have an influence on whether an individual would comply to the different assignments a RSU-app offers. Namely, one’s level of FoMO, gender and motivation to use a RSU-app were studied as predictors of compliance to the app (RQ 2). This study tested the app ‘Moment’ and its ‘Phone Bootcamp’ course for two weeks.
All in all, the hypotheses in this study cannot be confirmed. However, this study did find several important results. Participants did have lower PSU levels after the two weeks of this experiment, but this holds for participants in both the app and no-app condition. This finding might be explained by the fact that all the participants had to check their average screen time before the experiment: the participants might have been made aware of their smartphone use and this may have changed their smartphone use in a positive way. According to Ronda, Van Assema and Brug (2001), people who are more aware of their (problematic) mental health, have a more positive intention to do something about their mental health problem. This is in line with Dute et al. (2016) who state that increased awareness facilitates motivation to help reduce their mental health problems. Thus, awareness about the amount of one’s smartphone use itself might have already changed their behavior (such as using one’s smartphone in a less problematic way). Whether raising awareness of potentially problematic smartphone use is enough to improve healthy smartphone usage patterns, however, needs to be studied in future research.
Moreover, the participants had to self-report about their PSU, which may have had an influence on the results. The tests about the time participants spent on their smartphone indicated that there were, in fact, no significant differences in the time participants spent on their smartphone before and after the experiment. As the time spent on their smartphone was assessed by letting participants report on their exact screen time, this result may be a more reliable one compared to self-reports of PSU. Moreover, the overall PSU score was moderate (with a mean score of 30 out of 60), which indicates that the participants in this study sample did not perceive their smartphone use as problematic. As the time one spends on their smartphone is part of one’s level of PSU, it can be suggested that they also did not spend a problematic amount of time on their smartphone (approximately 1500 minutes per week). However, to the author’s knowledge, no studies have been done to indicate how much time spend on smartphones is problematic, which might be a suggestion for future research. In this way, people with high PSU can be ‘diagnosed’ and can be helped to get rid of this problematic use. The average PSU level and the time participants had spent on their smartphone might have influenced the results, as the final goal of RSU-apps is to improve unhealthy smartphone use and the participants did not use their smartphone in an ‘unhealthy’ way. Therefore, future research should also use a sample of participants with high levels of PSU, to see if RSU-apps actually work for people with high PSU.
Another finding of this study was that participants had a lower well-being after the two weeks of this experiment. However, in this study it was assumed that their well-being would increase after using the app. Interestingly, well-being decreased over the two weeks in both conditions: the app condition and the no-app condition. The lower well-being of all participants can be explained by several other factors. According to Reis, Sheldon, Gable, Roscoe and Ryan (2000) there are three basic psychological needs that influence well-being on daily basis: autonomy, competence and relatedness. They indicated that individuals with
more satisfied levels of autonomy (the degree to which individuals feel initiators of their own behavior), competence (the degree to which individuals feel able to achieve goals) and relatedness (the degree to which individuals feel related to one another) had a higher level of well-being, compared to individuals with less satisfaction of their basic psychological needs.
However, this is dependent on the individual’s level: whereas one individual might have had a decrease in their psychological satisfaction of relatedness (e.g. using the RSU-app disabled people to use their phone to connect with others, which made them feel less related) or a decrease in their need for competence (e.g. the RSU-app made people feel that they could not achieve their goals, as they could not use their phone for everything) and others might have had a decrease in the satisfaction of their phycological need autonomy (e.g. the RSU-app made people feel they were not initiators of their own behavior, as they could not use their phone whenever they wanted). However, the satisfaction of the basic psychological needs could also have been influenced by other factors, as the participants in the no-app condition also had a lower well-being. Therefore, future research should study the relation between the use of RSU-apps and the satisfaction of each distinct basic psychological need and whether this may influence one’s well-being.
In addition, daily positive or negative mood might have had an influence on the participants level of well-being (Reis et al., 2000). It might have been possible that the participants had a better mood in the pre-test than in the post-test, which may have been influencing the results. This could for example be explained by the fact that most of the participants were students and the second survey was distributed during the exam period. According to Sletta, Tyssen and Løvseth (2019), exam stress is related to a lower well-being of students. This is in line with Karademas (2007) who stated that one’s level of stress can be seen as a negative predictor of one’s well-being. This might also be an explanation for older participants, who might have had a stressful period on their work. Therefore, future research
should study the relation between RSU-apps and well-being taking into account other variables, such as mood and stress, that may predict well-being.
These factors may also have had an influence on compliance to RSU-apps. Namely, participants with a negative mood might have not been willing to use the RSU-app, or participants with high levels of stress might have not want to make time to comply to the app assignments. Other factors that influence compliance of young adults to health-related smartphone apps are, for example, the amount of effort they must put in to the app or the degree to which it was pleasant to use the app (Dennison, Morrison, Conway & Yardley, 2013). Future research should take these variables into account, when studying the compliance to RSU-apps.
Furthermore, it was tested whether the relationship between compliance to the RSU app and well-being was mediated by the time participants spent on their smartphones or PSU. Both mediations turned out to be insignificant. However, as suggested before, this might be related to the low PSU levels and the low time participants had spent on their smartphones. Moreover, it was also tested whether compliance of the participants to the RSU-app assignments, predicted lower levels of PSU, less time the participants spend on their smartphone and higher well-being. However, this was not the case. This finding might be explained by the fact that the participants in the app condition, in general, did not fulfill many of the assignments (on average 4 assignments out of 14). According to Payne et al. (2015) some users of mental health apps do not like being pressured by apps to perform a certain behavior and, therefore, see this pressure as a barrier to use the apps. Hence, it can be suggested that participants did not like to be pressured to use the RSU-app and performed less of the assignments.
This study also took into account the predictors of whether the participants would comply to the RSU-app assignments. The results indicated that FoMO, gender (females),
social purposes of smartphone use, intrinsic motivation and extrinsic motivation were no predictors for compliance to the RSU-app assignments. These findings could be explained in several ways. First, the participants did have high levels of offline FoMO (with a score of 18 out of 21) but a moderate score of online FoMO (36 out of 49). This might be explained by the fact that the participants could use other means to satisfy their online FoMO: if they were not allowed to use their phone, they still could use their laptops to go online and not miss anything of the online world. In addition, the results made clear that most of the participants did not comply to the assignments to a large extent, which may have been causing that this study did not find a relation between compliance to the RSU-app and FoMO. Future research should study the relation between FoMO and RSU-apps in a situation where the participants cannot use other means to satisfy their online FoMO and where the participants actually follow to RSU-app assignments.
Moreover, the other findings might be explained by the fact that the participants did not use their smartphones for a large amount for social purposes (with a maximum score of 3 out of 5) and were not highly intrinsically or extrinsically motivated (only one participant indicated that she was highly intrinsically motivated) to use the RSU-app. This last finding might be explained by the fact that the participants did not made the decision themselves to use a RSU-app. Normally, people only use RSU-apps when they are motivated to reduce their smartphone use and for this study the participants were forced to use the app. As the participants self-reported that they did not think they had high levels of PSU, this might have been leading to a low intrinsic motivation to actually use the RSU-app. Intrinsically motivated individuals are, as mentioned before, more successful in performing preventive health behavior (Sundar, Bellur & Jia, 2012), and as the participants were not highly intrinsically motivated to use the app, this might have had an influence on their use and compliance to the
app. Therefore, this can be seen as one of the limitations of this study and future research should study the effects of RSU-apps with more highly intrinsically motivated individuals.
This study also had some other general limitations. First, the sample for this study was gathered by using a convenience sample in combination with a snowball sample. This strategy saves time and money but is at the same time bad for the external validity, because this emerging adult group is not representative for all emerging adults. Thus, the findings of this study cannot be generalized. In addition, the greatest part of the sample consisted of people with a high education level and were from Dutch origin. Hence, it might be interesting for future research to test this topic with a more diverse sample. In addition, the aim was to gather 40 participants per condition, but this unfortunately did not work out as many participants did not fill out a real email address. This could have an influence on the outcome of this study, and therefore, future research can investigate the same topic, but with more participants.
Another limitation was that during the study it was noticed that only participants owning an iPhone could make use of the free code to unlock the ‘Moment coach’, as this code uses a shake function which did not work on Android phones. Therefore, the participants in the experiment condition were all iPhone owners, which had an influence on the number of participants in this condition. This might also have had an influence on the outcome of this study, thus future research can focus on people with both Android and iPhone devices. Furthermore, as mentioned before, this study was based on self-report and thus relies on the honesty of the participants. Unfortunately, this cannot be checked, and this might have had an influence on the results. Therefore, future research can investigate the influence of RSU-apps with another study design, such as an experiment based on data that was gathered by an RSU-app itself instead of participants. Also, longitudinal research is needed to test the long-term effects of RSU-apps on well-being. In addition, qualitative research can be done to investigate the reasons why participants use or comply to a RSU-app. Finally, future research can test
