Previous research has shown the capacity for mindfulness to be strongly associated with psychological well-being, that components of mindfulness show significant growth through young adulthood, and that this developing, malleable capacity is vital as individuals learn to deal appropriately with negative thoughts and unwelcome emotions. Smartphones, typically used in an automatic or experientially avoidant way, can undermine this development, leading to a decreased capacity for mindfulness. The purpose of these studies were to examine the extent to which smartphone use is negatively associated with young adults’ mindfulness and the degree to which increased cognitive and behavioral involvement with smartphones may exacerbate this relation using a newly developed conceptual model.
Method
Study 1 was conducted using self-report measures of mindfulness among a cross-sectional sample of university students aged 18–20 years (n = 668). Study 2 augmented Study 1 using objective measures of smartphone screen time and the cognitive regulatory components of mindfulness in a planned missingness design.
Results
Results indicate smartphone involvement (a compulsive pattern of use and cognitive preoccupation with one’s smartphone) to be significantly associated with lower trait mindfulness. Additionally, exploratory analysis of smartphone involvement as a mediator of the effect of smartphone use on mindfulness demonstrated a significant estimated indirect effect.
Conclusion
These results provide preliminary empirical support for the newly proposed conceptual model which posits associations between mindfulness and the use of smartphones in a cognitively and behaviorally involved way.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Access to smartphones for American youth has more than doubled in recent years, now reaching approximately 89% (Rideout & Robb, 2018). There is growing evidence that higher levels of smartphone use are associated with negative outcomes, including higher levels of anxiety and lower levels of self-esteem (Elhai et al., 2019), increased sleep difficulties (Liu et al., 2019; Twenge et al., 2017), problems with sustained attention and learning in the classroom (Ra et al., 2018), and greater risk of vehicle accidents (Centers for Disease Control & Prevention, 2024; McDonald et al., 2019; Ortiz et al., 2018). We posit that smartphone design lends itself to highly involved smartphone use, which is automatic, mindless, and used for the purpose of experiential avoidance, and that such highly involved use can negatively impact the capacity for mindfulness. Mindfulness is characterized as both an inherent, malleable capability and a state (Brown et al., 2007). Therefore, just as much as mindfulness training (i.e., practice engaging the state of mindfulness) has been shown to increase the inherent capacity for mindfulness, what Brown et al. (2007) referred to as trait mindfulness, the use of smartphones in a cognitively and behaviorally involved way could decrease it, with young people particularly vulnerable to this effect. While the development of mindfulness remains understudied, trait mindfulness has been associated with important indicators of psychological well-being (Flett et al., 2020; Semple & Burke, 2019). In contrast, experiential avoidance, or the refusal to attend to or acknowledge a thought or emotion that arises (Brown et al., 2007; Cardaciotto et. al., 2008) can be highly detrimental on long-term health and well-being (Kashdan et al., 2010; Mellick et al., 2019; Spinhoven et al., 2014; Tyndall et al., 2018). Therefore, an empirical test of relations among mindfulness, smartphone use, and smartphone involvement is needed to address these critical concerns.
The two current studies presented herein integrate and build on the multiple conceptualizations of mindfulness (Baer et al., 2006; Bishop et al., 2004; Shapiro et al., 2006) of which acceptance and attention are core components. As Bishop et al. (2004) described, mindfulness is composed of self-regulation of attention and an accepting awareness. Our working definition is adapted from Bishop et al. (2004) but diverges slightly by characterizing mindfulness in a similar manner to Brown and Ryan (2003), who conceived of it as both a trait and a state, both of which vary intraindividually and interindividually. Therefore, mindfulness comprises both the capacity to invoke the inherent traits of self-regulation of attention and an accepting awareness, and the actual state of such at a given moment. This ability to engage mindfully is malleable, implying that capacity to sustain attention, return attention to focus if distracted, inhibit elaborative processing, and be accepting of the present moment can all vary over time and be highly dependent on environmental supports or barriers for its development.
Furthermore, Shapiro et al. (2006) described the process of reperceiving, or an individual’s ability to step back and relate objectively to a current experience rather than getting caught up mentally and emotionally in elaborative judgment of the experience, which increases one’s ability to act mindfully rather than automatically. Self-report data (Khoury et al., 2013), objective measures of mindfulness components (Allen et al., 2012; Anderson et al., 2007; Wenk-Sormaz, 2005; Zanesco et al., 2013), and neurological studies (Grant et al., 2010; Hölzel et al., 2011; Lazar et al., 2005; Tang et al., 2010; Van Veen & Carter, 2002) all corroborate that mindfulness is plastic and that a purposeful increase in trait mindfulness through consistent behaviors and environmental supports (e.g., intervention) is possible. It follows that trait mindfulness could also be decreased by certain behaviors, and experiential avoidance and automatic reactivity are two potential mechanisms by which this could occur. Attempts at suppressing or avoiding aversive thoughts, feelings, or experiences are the opposite of mindful engagement in the present moment. Rather than sustaining attention on the aversive thought or feeling, the “experiential avoider” suppresses their awareness and redirects attention elsewhere. If reacting to unpleasant thoughts and feelings in this way occurs regularly, experiential avoidance can become an automatic reaction. The regular practice of engaging in an accepting awareness of the moment through self-regulation of attention allows for reperceiving (Shapiro et al., 2006), which increases one’s capacity for deautomization (i.e., a reduction of automatic responding; Deikman, 1963). Therefore, regularly avoiding awareness and reacting automatically could make engaging this capacity more difficult, especially under stressful circumstances.
Amid rapid development and changes within the brain, youth may be more vulnerable to environmental and social-behavioral factors that can decrease trait mindfulness, with a number of studies demonstrating that aspects of mindfulness are still developing during this time. Galla et al. (2020) were one of the first to conduct a longitudinal study explicitly tracking the natural development of mindfulness during adolescence. By examining nonreactivity in a sample of adolescents transitioning from middle school to high school, findings suggest that emotional nonreactivity to difficult experiences matures during adolescence, which was associated with changes in the cognitive control system. In a review of brain imaging studies, Blakemore and Choudhury (2006) observed that there are more changes occurring in the PFC, implicated in attention switching and response inhibition, than other parts of the brain throughout adolescence. Increased changes during adolescence in synaptogenesis (Huttenlocher & Dabholkar, 1997), gray matter reduction (Sowell et al., 1999), and myelination (Giedd et al., 1999) suggest that the capacity for mindfulness may still be developing. Both Booth et al. (2003) and Luna et al. (2010) have corroborated this by finding developmental differences in the PFC and performance differences in attention switching and response inhibition tasks between mid-adolescents and young adults. Given pruning and neural growth are happening at high rates during mid-adolescence (Thompson et al., 2000), this period of development is optimal for mental training to promote greater attentional control (Chin et al., 2021; Wass et al., 2012). If these capacities are still developing into adulthood and amenable to mental training, then it stands to reason that physiologically, they could also be especially vulnerable to behaviors that may decrease the capacity for mindfulness, such as regular smartphone use.
Smartphones can provide users with consistent, salient, and instant rewards, which then reinforce more usage (Deng et al., 2021). When an action is consistently rewarded, it begins to be automatically triggered with an expectation of subsequent reward (Neal et al., 2006). Over time, such a repeatedly triggered behavior can become a habit. The immediately accessible and highly salient rewards offered by smartphones, such as access to social networks, communication, news, and other online content, promote such automated behavior (Fullwood et al., 2017; Harwood et al., 2014). An increase in this kind of habit implies an increase not just in use, but also the level of involvement. In the era before the ubiquity of smartphones, researchers argued that, while mobile phone use had a negative impact, the degree of individuals’ cognitive and behavioral involvement with their phones was a key predictor of the extent to which use has deleterious effects on functioning (Walsh et al., 2010). In their survey of Australian youth aged 15–24 years, 85% reported moderate-to-highly involved use. Involvement, which was only moderately correlated (r = 0.30) with use, captures how people use their phones and the ways in which the devices impact their cognition and behavior, even when they are not actively being used. Recently, scholars have likened highly involved, problematic use to addiction, defining it as “a compulsive pattern of smartphone usage which can result in negative consequences that impair the daily functioning of the user” (Busch & McCarthy, 2021, p. 2). It has been found that highly involved smartphone users exhibit such negative symptoms as maladaptive dependency (Chen et al., 2017), withdrawal (Horwood & Anglim, 2018), and hindered social relationships, productivity, and physical health (Shin & Dey, 2013), to name a few. Hartanto and Yang (2016) found that a cognitive preoccupation with smartphones could result in short-term deficits in attention switching and inhibitory control. They found that users who were separated from their smartphones experienced heightened anxiety, which mediated an adverse effect on these functions. Compounding this, the newer capabilities of smartphones that extend far beyond texting and calling (i.e., social media apps, internet access, and online games) have exponentially expanded their potential for use to avoid boredom or other aversive feelings (Ahmed, 2019; Busch & McCarthy, 2021). Importantly, developmental changes in the brain during youth and young adulthood, particularly in the processing of social information, have been linked to changes in social behavior and increased salience of social experiences and ties to peers (Nelson et al., 2005). This increased salience can lead to difficulties balancing competing social demands and meeting social expectations of constant connectivity (Rotondi & Tomasuolo, 2017). That is, friends and peers are highly influential (Smetana et al., 2006), and using phones to stay in contact functions to satisfy this need for increased social inclusion and connectedness (Walsh et al., 2010).
Together, the research reviewed above provides foundational support for the premise that smartphone use negatively impacts mindfulness, with this effect strengthening as cognitive and behavioral involvement with smartphones increases. Insofar as mindfulness training has been shown to increase the inherent capacity for mindfulness, we argue that the use of smartphones, especially in a cognitively and behaviorally involved way, could decrease it, with the still-developing capacity of youth being particularly vulnerable to this effect. The primary aim of the two present studies was the preliminary test of a conceptual model that describes theoretical connections between mindfulness, smartphone use, and cognitively and behaviorally involved smartphone use (aka smartphone involvement). Using cross-sectional survey data (Study 1) as well as objective measurement of smartphone use and the cognitive regulatory components of mindfulness (Study 2) among an undergraduate student sample aged 18–20 yeras, we hypothesized that (1) smartphone use will be positively correlated with smartphone involvement, (2) smartphone use will be a significantly and negatively associated with mindfulness, (3) smartphone involvement will be significantly and negatively associated with mindfulness, and (4) smartphone involvement will moderate the association between smartphone use and mindfulness, such that a higher level of smartphone involvement will be associated with a stronger negative association between mindfulness and smartphone use. These studies are the first empirical test of hypotheses derived from the conceptual model. Our model conceives of smartphone involvement acting as a moderator of the relation between smartphone use and mindfulness. There is, however, a plausible, alternate conception in which smartphone involvement acts as the mechanism, or mediator, of smartphone use’s effect on mindfulness. In light of this, we also conduct a post hoc, exploratory analysis that tests smartphone involvement as a mediator of this relation. The present studies are the first empirical test of the implications of these theorized connections.
It is expected that a better understanding of the relations among these variables will help illuminate the mechanisms by which smartphones could impact mindfulness among young adults. This understanding could then facilitate the design of methods aimed at reducing the potential negative impacts of smartphone use. In the present studies, we limit the focus to the relations among smartphone use, smartphone involvement, and mindfulness, due to the unknown nature of said relations. It is hoped that by elucidating these constructs, a clearer framework for studying their relations to psychological well-being and other indicators of positive adaptation can be discovered. For Study 1, hypotheses are tested using student self-report data. Study 2 incorporates objective measures to improve measurement precision and uses a planned missingness design allowing for these measures to be given to a subset of the larger sample used in Study 1 but retaining the power of the larger sample. This additional study augments the findings using self-reported data in Study 1 with objective measures collected for two out of the three latent variables (i.e., smart phone use and mindfulness).
Study 1 Method
Study 1 was an observational, cross-sectional study that tested the above hypotheses using self-report data from a large sample of undergraduates. We assessed the relations between smartphone use, smartphone involvement, and mindfulness, including an exploratory relation whereby smartphone involvement mediates the effect of smartphone use on mindfulness.
Participants
Participants in study 1 included 668 undergraduates aged 18–20 years recruited from the a public university in the Southeastern United States (see Table 1 for demographic characteristics). Recruitment took place in the Fall of 2015 semester through a posting on the university’s Department of Psychology Participant Pool website, postings on Facebook pages for undergraduates that attend the university, contacting professors of large psychology classes at the university and asking them to offer extra credit, and contacting deans of other schools at the university and asking them to forward the information to their students. Participants who were psychology majors received course credit for participation, and all participants who participated in the study were entered into a drawing to win a 16 GB Apple iPad Air.
Table 1
Demographic characteristics of Study 1 sample and Study 2 subsample
Characteristic
Study 1 number
Study 1 percentage
Study 2 number
Study 2 percentage
Age
18
304
45.5
21
38.2
19
209
31.3
8
14.5
20
155
23.2
26
47.3
Gender
Female
520
77.8
50
86.1
Male
148
22.2
5
13.9
Race/ethnicity
White, not Hispanic
545
81.6
40
72.7
African American, not Hispanic
55
8.2
8
14.5
Multiracial, not Hispanic
18
2.7
*
*
Asian or Pacific Islander
16
2.4
*
*
American Indian or Alaska Native
4
*
*
*
Hispanic
30
3.7
*
*
Study 1 n = 668; Study 2 n = 55
*Sample and subsample sizes of fewer than five individuals are masked to reduce the risk of deductive disclosure
Inclusion requirements were as follows: 18–20 years old; current student status at the university; have owned and used an Android smartphone or an iPhone version 5 or later for at least the previous 3 months; have normal or corrected-to-normal vision; and speak English fluently.
Procedure
We collected data through an online survey during the Fall of 2015, using SoGoSurvey (https://www.sogosurvey.com/). Before beginning the survey, participants were asked questions on the inclusion criteria (age, enrollment status, phone ownership, English-speaking, etc.) and a total of 95 individuals did not meet the inclusion criteria. We obtained electronic informed consent for Study 1 from each participant prior to beginning the online survey. After participants completed the measures, we asked for their consent to be contacted to participate in Study 2. A total of 633 out of the initial 668 participants (94.7%) gave consent to be contacted to participate in Study 2.
Measures
We asked all study participants to complete questions regarding inclusion criteria, a brief demographic questionnaire and consent to be contacted for Study 2, the Mindfulness Attention Awareness Scale (MAAS), questions about their smartphone use and ownership, an adapted version of the Mobile Phone Involvement Questionnaire (MPIQ), a set of supplementary questions about smartphone use compiled by the researcher, and a restatement of the previous question regarding daily smartphone use, in a different form, in order to check validity. In addition, we recorded the time required for each participant to complete the measures, in order to screen for participants who completed the measure much more quickly than others.
Mindfulness
We measured mindfulness using the MAAS (Brown & Ryan, 2003), a 15-item self-report scale designed to measure trait mindfulness. This scale focuses on measuring the capacity for attention to and awareness of the present moment. The items each have six response options ranging from 1 (almost always) to 6 (almost never), with higher scores reflecting more mindfulness. Examples of items are “I find it difficult to stay focused on what’s happening in the present,” and “It seems I am ‘running on automatic,’ without much awareness of what I’m doing.” The internal consistency (coefficient alpha) for their sample of 327 university students was 0.82. Subsequent studies using college samples corroborated the validity and factor structure of the MAAS (e.g., MacKillop & Anderson, 2007). Numerous studies have shown that participants involved in interventions designed to increase mindfulness show significant increases in MAAS scores versus active control (e.g., Cohen-Katz et al., 2005; Shapiro et al., 2007).
Smartphone Use
We asked participants at two separate points in the online survey to estimate as accurately as possible their daily smartphone usage, once by estimating total time of use per day and once by selecting from a list of ranges (an hour or less, between 1–3 hr , between 3–5 hr, between 5–7 hr, more than 7 hr). We noted that they consider all uses except listening to music (e.g., calling, texting, social media platforms, gaming, watching videos). We also asked that participants estimate how old they were when they first owned and regularly used a smartphone. Two participants were excluded from analysis because the estimated time of usage given in their first response did not fall within ± one range of their second response (e.g., one excluded respondent first reported using their phone 8 hr a day then chose 3–5 hr when asked the second time). The estimated length of time that participants owned a smartphone was included in the analyses as a covariate.
Smartphone Involvement
We measured smartphone involvement using a modified version of the Mobile Phone Involvement Questionnaire (MPIQ) and a supplemental questionnaire developed by the authors. Walsh et al. (2010) developed the MPIQ to measure cognitive and behavioral involvement with phones. The MPIQ is an 8-item measure of the cognitive energy people devote to their phones even when not using them; the extent of their automatic, mindless use; and use that negatively impacts other activities. The item responses are scored on a 7-point ordinal scale, 1 (strongly disagree) to 7 (strongly agree). Because we were exclusively interested in smartphone use, “mobile phone” was changed to “smartphone” in all items. Examples of items are “I often think about my smartphone when I am not using it,” and “I interrupt whatever else I am doing when I am contacted on my smartphone.” Initial validation using a sample of Australians aged 16–24 years showed a single-factor solution with moderate reliability (α = 0.78). Correlation with frequency of phone use was significant (r = 0.37, p < 0.001), suggesting that phone involvement is a distinct construct from phone use.
Again, we conceive of smartphone involvement as a continuous construct, with users falling along a continuum. The developers of the MPIQ (Walsh et al., 2010) initially reported their results as dichotomous, with respondents designated as highly involved if they passed an arbitrary threshold. Because of this, we altered the responses from a scale measuring agreement with the statement (strongly disagree to strongly agree) to responses that allow for reporting of the frequency of engaging in activities described by the items (almost never to almost always) as a way to better measure responses along the continuum. These responses were on a 5-point scale rather than the originally used 7-point scale, as five distinct response anchors were conceptually identified as meaningful.
The MPIQ addresses, in part, the type of involved use that we theorize could contribute to decreased mindfulness. It has items related to using the phone “for no particular reason,” “los(ing) track of how much” the phone is used, and interrupting other activities if an alert is received (Walsh et al., 2010). These items address automatic, compulsive use, but not use for the purpose of avoiding uncomfortable feelings and/or thoughts (i.e., experiential avoidance). Excluding the latter would give an incomplete assessment of an individual’s level of cognitive and behavioral involvement. As described in the introduction, the MPIQ was developed before the ubiquity of smartphones and at a time when the technology of the phones, the networks, and the apps were not nearly as advanced. Today’s smartphone user has easier and more available access to vastly more content and reward for potential experiential avoidance than at the time of the initial study. This necessitates the inclusion of items designed to capture this type of use. Therefore, we formed a pool of 7 items inquiring about the frequency of experientially avoidant smartphone use and use in potentially dangerous situations, also based on a 5-point ordinal scale for consistency with the MPIQ items. These items were based on aspects of the type of highly involved use that could negatively impact mindfulness that are not adequately covered on the MPIQ and other validated scales measuring problematic smartphone use (e.g., Mobile Phone Problem Use Scale (MPPUS; Bianchi & Phillips, 2005); Problematic Mobile Phone Use Questionnaire (PMPUQ; Billieux et al., 2008)). Several graduate students and PhD-level researchers assisted by evaluating the initial pool of questions for clarity, and some questions were rephrased or edited. Examples of items are “I use my smartphone when I feel awkward or uncomfortable,” and “I use my phone while I’m walking for texting, email, reading, or social media.”
Data Analyses
All data analyses were conducted using Mplus v7.4 (Muthén & Muthén, 2015). We used confirmatory factor analysis (CFA), a hypothesis-driven form of factor analysis, in order to test the factor structures of the latent variables of smartphone involvement and mindfulness, using maximum likelihood estimation with robust standard errors (MLR). We chose CFA to test whether the observed data were consistent with the measurement model hypothesized based on the previous research and theory cited above. There were no missing data in study 1. All confidence intervals reported herein were derived using percentile bootstrapping with 3000 draws. Bootstrapping was used in order to estimate potentially asymmetric confidence intervals, reflecting the fact that products of coefficients (indirect effects) have asymmetric sampling distributions. Bootstrapped confidence intervals were not calculated for the hypothesized moderation effects because the normally distributed sampling distribution allows for accurate calculation of confidence intervals using the standard error. Based on parameter estimates and relationships from the theoretical model, we performed a power analysis using G*Power (release 3.1.9.2; Faul, et al., 2007). Analyses indicated that the present study had statistical power (1-β) of 0.80 to detect an incremental R2 of 0.012 (relative to an R2 of 0.5 for the restricted model) for the hypothesized moderation effect with n = 668. Because it was treated as an interaction effect, the moderation effect was the most difficult to detect. Power to detect main effects was higher.
Measurement Model
The first step in analysis was evaluating the fit of our study 1 measurement model. This model included the adapted MPIQ and supplementary questions to measure smartphone involvement and the MAAS to measure self-report mindfulness. To assess reliability, we calculated coefficient omega for the MAAS and our smartphone involvement measure. We chose to use coefficient omega rather than alpha because omega has less restrictive assumptions (alpha assumes tau-equivalence), does not increase with mere scale length, is more sensitive to multidimensionality, and allows for generation of confidence intervals.
Model Fit
We assessed model fit for our theorized structural model using multiple goodness-of-fit statistics: chi-square test of fit (χ2), Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA), and Standardized Root Mean Residual (SRMR). Hu and Bentler (1999) recommend the following thresholds to indicate good model fit: CFI > 0.95, TLI > 0.95, RMSEA < 0.06, and SRMR < 0.08.
Hypothesis Testing
For the sake of clarity, we tested Hypotheses 1–3 in the context of a single structural equation model (Fig. 1a) including years of smartphone ownership as a predictor of mindfulness, smartphone use, and smartphone involvement. Because this model is structurally saturated using the measurement model established in the CFA, then acceptable fit for that model also indicates acceptable fit for this model. To test the moderation Hypothesis 4, we used the Mplus function XWITH to create a latent product term from the latent smartphone use variable and the latent smartphone involvement variable. XWITH uses a latent moderated structural equations (LMS) method to estimate multiple latent interactions, which has been shown to provide efficient parameter estimates and unbiased standard errors (Klein & Moosbrugger, 2000).
Fig. 1
A Structural model for Study 1 test of Hypotheses 1, 2, and 3. Correlation and unstandardized factor loadings (standard errors) shown. Disturbances not shown. Years of ownership is entered as a covariate. B Structural model for study 1 Hypothesis 4a (moderation). Unstandardized factor loadings (standard errors) shown. Disturbances not shown. Years of ownership is entered as a covariate. C Structural model, study 1 mediation. Indirect effect was significant (a × b) = − 0.02, [BC] 95% CI [− 0.03, − 0.01]. Unstandardized factor loadings (standard errors) shown. Disturbances not shown. Years of ownership is entered as a covariate
×
Study 1 Results
Descriptive Data
Descriptive statistics for smartphone ownership and use are found in Table 2, with bivariate correlations shown in Table 3. The age reported for first owning a smartphone ranged from 10 to 19 (M = 14.4, SD = 1.7), with a mean duration of smartphone ownership of 4.4 years (SD = 1.7). Finally, in estimating their daily smartphone use, our sample reported a range of 0.5 to 18 hr of use per day (M = 5.7, SD = 3.5). Two participants were excluded from analysis because there was a discrepancy in their estimated time of usage on items used to check validity of responding.
Table 2
Descriptive statistics for smartphone ownership and use
Min
Max
Mean
SD
Median
Age first smartphone ownership (years)
10
19
14.4
1.7
14
Study 1 (minus Study 2 sample, n = 618)
10
19
14.3
1.7
14
Study 2 (n = 55)
10
18
14.8
1.9
15
Duration smartphone ownership (years)
< 1
8
4.4
1.7
4
Study 1 (minus Study 2 sample, n = 618)
< 1
8
4.4
1.7
4
Study 2 (N = 55)
2
8
4.3
1.8
4
Self-reported smartphone use (hours/day)
0.5
18
5.7
3.5
5
Study 1 (minus Study 2 sample, n = 618)
0.5
18
5.6
3.4
5
Study 2 (n = 55)
1
15
6.3
3.4
6
N = 668
Table 3
Study 1 bivariate correlations
1
2
3
4
5
1. SUSELF
1.00
2. YOWN
0.11*
1.00
3. SI
0.37*
0.24*
1.00
4. Mindfulness
0.18*
0.08
0.49*
1.00
5. SUSI
0.31*
0.20*
0.84*
0.43*
1.00
N = 668. SUSELF self-reported smartphone use, YOWN self-reported years of smartphone ownership, SI smartphone involvement, SUSI latent interaction term for smartphone use and smartphone involvement
*p < 0.05
For the ordinal items, we also examined the distributions for possible strong ceiling or floor effects. We found asymmetry among some of these responses, with four of the smartphone involvement and two of the mindfulness items having modes at an extreme. Due to the significant computational difficulties inherent in the alternative approach of treating this large number (30) of indicators as ordinal with our moderate sample size, ordinal items were treated as continuous.
Measurement Model
Both the MAAS (ω = 0.82, bootstrap corrected [BC] 95% CI [0.80, 85]) and the hybrid smartphone involvement measure (ω = 0.90, [BC] 95% CI [0.89, 91]) showed adequate reliability.
Evaluating Model Fit
All items loaded significantly onto their respective factors (loadings ranging from 0.37 to 0.72 on the MAAS and 0.41 to 0.69 on the hybrid SI scale—see Table S1 in the Supplementary Information Section). The chi-square value for the overall model fit was significant (χ2(426) = 874.90, p < 0.001), suggesting a lack of fit between the hypothesized model and the data. However, due to the sensitivity of χ2 to small degrees of mis-fit, other fit indices are emphasized (Kline, 2010). For the other fit statistics, Hu and Bentler (1999) recommend the following thresholds to indicate good model fit: CFI > 0.95, TLI > 0.95, RMSEA < 0.06, and SRMR < 0.08. Examination of these indices showed mixed results regarding model fit, with CFI = 0.93, TLI = 0.92, RMSEA = 0.039, 95% CI [0.035, 0.042], and SRMR = 0.043. While the RMSEA and SRMR meet the recommended thresholds, the CFI and TLI values are slightly below the recommended cutoffs. These values both depend on the size of the correlations in the measured variables, which in the case of this dataset low, particularly on the MAAS. Rather than attempting post hoc modification of this established measure, we will accept marginal fit and interpret our estimates with caution. We hypothesized that the MPIQ and these questions together measure a unidimensional factor of smartphone.
Hypothesis Testing
Two of the three bivariate hypotheses (Hypotheses 1–3) were supported. Regarding Hypothesis 1, there was a significant positive correlation between the latent variables smartphone use and smartphone involvement, r = 0.37, p < 0.001. Hypothesis 2, which stated that there would be a significant and negative relation between smartphone use and mindfulness, was not supported (b = 0.00, [BC] 95% CI [− 0.01, 0.01], z = 0.18, p = 0.861). Hypothesis 3, which posited smartphone involvement as a significant negative predictor of mindfulness, was supported (b = − 0.43, [BC] 95% CI [− 0.58, − 0.31], z = 5.99, p < 0.001). The coefficient for mindfulness regressed on this latent product term in the structural equation model was not significant (b = 0.01, 95% CI [− 0.01, 0.03], z = 0.72, p = 0.48), indicating no support for Hypothesis 4 (smartphone involvement as a moderator; Fig. 1b).
Exploratory Analysis
Although it was hypothesized that smartphone involvement is a moderator of the effect of smartphone use on mindfulness, it is also plausible that smartphone involvement, instead of moderating this effect, is a mechanism by which the effect occurs. Therefore, we also explored this potential mediation, testing the model shown in Fig. 1c. Examination of fit indices again showed mixed results regarding model fit, χ2(484) = 1005.19, p < 0.001, CFI = 0.92, TLI = 0.91, RMSEA = 0.040, 95% CI [0.037, 0.044], SRMR = 0.043. We found a statistically significant product of coefficients (a*b = − 0.02, [BC] 95% CI [− 0.03, − 0.01]), indicating support for the possible indirect effect.
Study 2 Method
In Study 2, we again assessed the relations between smartphone use, smartphone involvement, and mindfulness, as we did in study 1, but also included objective measures of two of the three latent variables (smartphone use and mindfulness) to augment the strictly self-report data used in Study 1. Since having multiple measures of latent constructs allows for better modeling of the error structures of said measures (e.g., Kaplan et al., 1988), we tested our hypotheses in Study 2 using the two-method, planned missingness design as described in Graham et al. (2006). In this design, “inexpensive” (in terms of money, resources, and/or time) measures are given to the full sample (Study 1), while more “expensive” measures are used on a portion of respondents to facilitate more precise measurement (Study 2). A planned missingness context allows for more power than with the expensive measures alone and better construct measurement than with the inexpensive measures alone (Graham et al., 2006). This is accomplished because the more expensive (and more valid) measure is used to help model the response bias of the cheaper measures. With mindfulness, we accomplished this by using a second-order factor model, in which items on the MAAS load onto a single MAAS factor, becoming the third indicator of mindfulness, along with the two expensive measures. Because there was only one self-report (i.e., inexpensive) indicator of smartphone use, we could not explicitly model the response bias. However, having the second, expensive indicator improved the measurement of this latent variable. This method of handling missing data is generally referred to as the gold standard (Schafer & Graham, 2002) in the field and is valid provided the missing at random (MAR) assumption is met (Enders & Baraldi, 2018), which can be assumed in our study given our use of a random number generator in SAS 9.4.
In addition to Hypotheses 1–3, tested here in the context of a single structural equation model including all covariates as predictors of mindfulness, smartphone use, and smartphone involvement (Fig. 2a), Study 2 also assessed an exploratory relationship whereby smartphone involvement mediates the effect of smartphone use on mindfulness.
Fig. 2
A Structural model for Study 2 Hypotheses 1, 2, and 3. Unstandardized factor loadings (standard errors) and correlation shown. Disturbances not shown. Years of ownership is entered as a covariate. B Structural model for Study 2 Hypothesis 4b (moderation). Unstandardized factor loadings (standard errors) shown. Disturbances not shown. Years of ownership is entered as a covariate. C Structural model for testing Study 2 mediation. Indirect effect was significant (a × b) = − 0.25, [BC] 95% CI [− 0.90, − 0.05]. Unstandardized factor loadings (standard errors) shown. Disturbances not shown. Years of ownership is entered as a covariate
×
Participants
Beginning with the first eight participants in Study 1, we numbered each group of eight consecutive participants 1 to 8, in order of completion and used SAS 9.4 (SAS Institute, Cary NC) to generate a random number from 1 to 8 for selection. In total, we invited 350 study 1 participants to take part in Study 2, with 214 (61.1%) not responding, 81 (23.1%) declining to participate, and 56 (16.0%) agreeing to participate, of whom one did not meet their appointment for Study 2 and did not respond to follow-up contact. Of the 55 participants from Study 1 who participated in Study 2, our compliance rate was 100% (see Table 1 for demographics).
Procedure
We invited the participants who agreed to take part in Study 2 to come into the laboratory, where we then asked them to give informed consent for participation in two phases of Study 2. This consent form addressed the two phases of Study 2 separately and allowed participants to consent to each phase separately. In the first part of Study 2, participants completed two cognitive assessments that measure attentional regulatory components of mindfulness on an iPad. These measures represent constructs (i.e., sustained attention, attention switching, and inhibition of elaborative processing of thoughts, feelings, and sensations) associated with the conceptualization of mindfulness by Bishop et al. (2004), which includes self-regulation of attention as an inherent component. The primary researcher administered the tasks in random order (decided by a coin flip).
At the completion the first part of Study 2, we asked those who agreed to participate in the second phase to download a free app to their smartphones (Moment for iPhones) that automatically records how long they use their phones each day without having to self-report. We asked participants to enable Moment to run in the background, invisibly tracking the amount of time their phone is being actively used. A researcher then contacted participants every 8 days for the next 32 days by email and asked them to forward the number of minutes that they used their phones each day for the previous 8 days. The Moment app contains a feature that allowed participants to easily forward daily total usage.
Measures
In order to objectively measure sustained attention and attention switching, we used, respectively, the Sustained Attention Response Task (SART; Robertson et al., 1997) and the Trail Making Test (TMT; Army Individual Test Battery (AITB), 1944). We chose these measures because they allow for valid measurement of these aspects of mindfulness with a minimal time investment from the participants and could be administered in a uniform manner on an iPad.
Sustained Attention
The SART (Robertson et al., 1997) is a continuous performance paradigm requiring response to frequently presented non-targets and a withholding of the response for occasional targets. Participants are shown 225 single digits for 250 ms each, with a 900-ms mask, and are asked to tap the iPad screen in response to every digit other than “3.” If a “3” appears, they are instructed to not respond. This task is designed to require continuous attention to response, be sensitive to brief lapses in attention, and have minimal demands on other cognitive processes (e.g., memory, planning, and general cognitive effort). The SART places demands on maintaining attentional focus in two ways: by having long and unpredictable intervals between targets and requiring continuous performance over the duration of the 225-trial/4.3-min task. Testing by Robertson et al. (1997) showed SART performance predicted self and informant reports of everyday attentional failures. Further study has demonstrated that the SART measures sustained attention performance, which includes but is not limited to response inhibition (Manly et al., 1999). We used the number of omission errors on the SART as the indicator for the latent variable of sustained attention.
Attention Switching
To objectively measure attention-switching ability, we used the two-part Trail Making Test (TMT; AITB, 1944). This test was initially published as part of the Army Individual Test Battery (1944) as a paper-and-pencil test but has been adapted for use with the touchscreen of the iPad. Part A (TMT-A) requires connecting numbers 1–25 with a single line as quickly as possible while still maintaining accuracy, while part B (TMT-B) requires drawing a similar line connecting alternating numbers and letters in order (i.e., 1-A-2-B-3-C etc.). The score recorded for each part is the time required to complete each “trail,” or connecting line. A recent comprehensive review and validation of the TMT provides strong support to the initial assumptions, that attentional switching ability is the primary variable accounting for variance in B-A difference scores (Sánchez-Cubillo et al., 2009). Mayr and Keele (2000) demonstrated that the shifting from one task-set to the other requires not just a shift of attentional focus, but also inhibition of attention to the currently irrelevant task-set, indicating that B-A difference scores provide a relatively clear measure of the capacity for attention switching.
Smartphone Use
The variable of interest in this study, minutes of smartphone use per day, was measured using the Moment app. A previous study by Lee et al. (2014) used a similar, self-developed app to unobtrusively monitor smartphone use. They found significantly different levels of use in the first 2 days of monitoring as compared with the rest of the study, so they excluded those days. Accordingly, we also excluded the first 2 days of monitoring.
Data Analyses
All data analyses were conducted using the software and methods described in detail in the “Data Analyses” section of Study 1.
Study 2 Results
Descriptive Data
Descriptive statistics for Study 2 measures are found in Table 4 and correlations are shown in Table 5. Additionally, we examined the objectively measured hours of smartphone use, finding a skew of 0.20. Based on previous literature, there was no reason to expect non-normality on either the SART or TMT, but we examined this data visually in order to screen for outliers and calculated skew for each to be 1.98 and 1.58, respectively.
Table 4
Descriptive statistics for Study 2 measures
Min
Max
Mean
SD
Median
Smartphone use (mean hours/day)
0.75
7.35
3.83
1.31
3.73
TMT B-A (s)
6.9
42.1
23.3
9.3
21.7
SART omissions (no. of missed targets)
0
7
1.02
1.38
1
n = 55. Smartphone use measured by Moment smartphone application. TMT B-A difference in secs between TMTA and TMTB completion
Table 5
Study 2 bivariate correlations
1
2
3
4
1. SU
1.00
2. YOWN
0.19*
1.00
3. SI
0.62*
0.24*
1.00
4. Mindfulness
0.39*
0.11
0.64*
1.00
n = 55. SUSELF self-reported smartphone use, YOWN self-reported years of smartphone ownership, SI smartphone involvement
*p < 0.05
Planned Missingness Design
Among participants in Study 2, correlations between the inexpensive and expensive measures ranged from moderate to near zero. Smartphone use as measured by the Moment app showed a moderate correlation with self-reported smartphone use, r = 0.38, p < 0.001. Our participants tended to overestimate their smartphone use, reporting in Study 1 that they used their phones an average of 6.4 hr a day (SD = 3.4), while the Moment app found that participants averaged 3.8 hr of use each day (SD = 1.3). The two objective measures of mindfulness did not significantly correlate with MAAS (r = 0.01, p = 0.508 for the TMT B-A score and r = 0.15, p = 0.207 for SART omissions). The TMT B-A score was also not significantly correlated with SART omissions, r = 0.24, p = 0.081.
Measurement Model
The first step in analysis was evaluating the fit of our Study 2 measurement model to see how well the latent variables of smartphone involvement and mindfulness are being measured via self-report. All of the following analyses were also conducted using Mplus v7.4 (Muthén & Muthén, 2015).
Evaluating Model Fit
We used confirmatory factor analysis (CFA), a hypothesis-driven form of factor analysis, in order to test the factor structures of smartphone involvement and mindfulness using maximum likelihood estimation with robust standard errors (MLR). All items loaded significantly onto their respective factors aside from the TMT B-A score (b = 0.87, SE = 1.45, z = 0.60, p = 0.549). Due to the small standardized loading (0.09) of TMT B-A on mindfulness, this measure was removed post hoc from our CFA and hypothesis testing. Fit indices of the revised measurement model (Fig. 2a) showed mixed results regarding model fit: χ2(517) = 1056.99, p < 0.001, CFI = 0.92, TLI = 0.91, RMSEA = 0.038, 95%CI [0.035, 0.041], SRMR = 0.079. As with the Study 1 CFA, the RMSEA and SRMR meet the recommended thresholds, while the CFI and TLI values are below the recommended cutoffs. We again chose to accept marginal fit and interpret our estimates with caution, rather than attempting post hoc modification of this established measure.
Hypothesis Testing
Our results demonstrated support for two of the three bivariate hypotheses. Regarding Hypothesis 1, there was a significant positive correlation between the latent variables smartphone use and smartphone involvement, r = 0.61, p < 0.001. Hypothesis 2, which stated that smartphone use would have a significant negative relation with mindfulness, was not supported (b = 0.03, [BC] 95% CI [− 0.51, 1.13], z = 0.12, p = 0.90). There was support for Hypothesis 3, which posited smartphone involvement as a significant predictor of reduced mindfulness (b = − 0.83, [BC] 95% CI [− 1.97, − 0.51], z = 4.30, p < 0.001). To test the moderation hypothesis (Hypothesis 4), we again used the Mplus function XWITH to create a latent product term from the latent smartphone use variable and the latent smartphone involvement variable in the model shown in Fig. 2b. The coefficient for mindfulness regressed on this latent product term in the structural equation model was not significant (b = − 0.42, 95% CI [− 1.08, 0.37], z = 1.04, p = 0.30), indicating no support for Hypothesis 4.
Exploratory Analysis
As with Study 1, we also explored the potential mediation of the effect of smartphone use on mindfulness by smartphone involvement. Examination of fit indices indicated marginal model fit: χ2(548) = 1115.54, p < 0.001), CFI = 0.91, TLI = 0.91, RMSEA = 0.039, 90% CI [0.036, 0.043], SRMR = 0.080. We found a significant product of coefficients (a*b = − 0.25, [BC] 95% CI [− 0.70, − 0.05], Fig. 2c), indicating support for the possible indirect effects.
Discussion
The present study was designed to test our theoretical model, which hypothesizes relations among smartphone use, smartphone involvement, and mindfulness, in a sample of undergraduates aged 18–20 years. Importantly, the findings were essentially the same for Study 1, using self-report data only, as for Study 2, which added objective data in a planned missingness design in order to improve the measurement of our latent constructs. The main findings supported the hypothesis that the use of smartphones in a behaviorally and cognitively involved manner is significantly associated with lower levels of trait mindfulness. In addition, findings from exploratory analyses suggested that this behavioral and cognitive involvement could explain the relation between smartphone use and mindfulness. This supports the idea that the manner in which young adults use their smartphones is more important than how much they use them, in relation to trait mindfulness. These results are an important, novel link between the positive outcomes associated with mindfulness and the use of these ubiquitous devices among young adults. While a great deal of research has been devoted to increasing mindfulness and its correlates, increasing our understanding of the impact of smartphones on mindfulness may help identify relevant areas for interventions to reduce any deleterious effects of smartphone use, particularly among youth.
While our understanding of the impact of smartphones on developing capacities for attention and awareness is limited, there exists a small but growing foundation of studies examining smartphone use and mindfulness. The present results are consistent with past studies that demonstrated that signs of high smartphone involvement (e.g., texting while driving, anxiety when separated from phone) are associated with lower levels of present moment awareness and lower capacity for sustained attention (Ra et al., 2018), lower trait mindfulness (Feldman et al., 2011), decreased abilities for attention switching and distraction inhibition (Hartanto & Yang, 2016), and more distractibility while driving (Centers for Disease Control & Prevention, 2024; McDonald et al., 2019; Ortiz et al., 2018). Our results were also consistent with Walsh et al. (2010), who found smartphone use and smartphone involvement to be moderately correlated but quantitatively distinct constructs.
Although the present study builds on previous literature on phone use and involvement, the rapid advancement of technology means that we studied a qualitatively different device than many previous studies. Much of the past research on the effects of phone use quantified use as frequency of texting and calling, in which recent research has shown to constitute less than half of the bulk of smartphone use (GfK MRI, 2016). Therefore, many of these previous studies may significantly underestimate the potential impact of smartphone use.
Another significant aspect of the present study is that we build on and improve on the aforementioned self-report-only studies by using the Moment smartphone app (or another similar app) to objectively measure smartphone use and the SART to objectively measure mindfulness. That we found only a moderate correlation between estimated daily use through self-report and daily use measured objectively highlights the importance of including such measures in smartphone research. Notably, participants in both studies tended to overestimate their daily smartphone use as compared with objectively measured use. The small correlation between the SART and the MAAS in our sample highlights the importance of continued refinement of our understanding of the construct of mindfulness and how to best measure it.
The present study extended prior work by demonstrating that cognitive and behavioral involvement is significantly associated with trait mindfulness, and by providing evidence of a possible mechanism of action for the proposed effect of smartphone use on mindfulness. Our hybrid smartphone involvement measure includes items related to automatic reactivity and experiential avoidance, which are essentially the opposite of mindful action and awareness. These processes can be compared to parallel but inverse processes that occur in mindfulness training where participants learn to act mindfully, with intent, rather than to react reflexively, and to allow themselves to experience uncomfortable thoughts, feelings, and experiences when appropriate. Therefore, it follows logically that in the same way purposefully practicing engaging in state mindfulness increases trait mindfulness, and consistent practice with avoiding state mindfulness would reduce this capacity. The results of our post hoc exploratory analyses provide the first empirical evidence that this may be the process by which smartphone use impacts mindfulness.
This study also extends the literature by integrating a measure of how long participants have been smartphone users. In both studies, duration of smartphone ownership was not significantly associated with mindfulness but was significantly associated with both smartphone use and smartphone involvement. This indicates that participants who have used a smartphone longer tend to use them more each day and also tend to use them in a more cognitively and behaviorally involved manner. The relation between duration of use and other study variables was not a focus of this study, but these findings are consistent with our argument that smartphone design elicits more use and more involved use, as levels of both smartphone use and involvement increased with increased duration of ownership.
Limitations and Future Directions
There are several limitations with the present study that should be noted. First, the use of a cross-sectional design limits our ability to make inferences about temporal causal relationships among variables. This is especially notable, as our conceptual model makes specific assertions regarding temporal relationships, namely that smartphone involvement can negatively impact mindfulness. While there is theoretical support for this assertion that is further supported by the findings of the current study, future prospective, longitudinal studies are necessary to test the direction of these relations, as experimental studies would not be possible with smartphone involvement. It is likely that such studies may find reciprocal effects between mindfulness and smartphone involvement. While our model focuses on the potential impact of smartphone involvement on mindfulness, it seems apparent that one’s limited capacity for mindfulness could be an indication of potential vulnerability to highly involved use. It will be important, therefore, to ascertain whether purposeful increases in mindfulness could decrease smartphone involvement and/or protect against its potential negative impacts.
Second, all of these results come from structural models that showed less-than-ideal fit. As mentioned, we chose to move forward and interpret results with marginal fit rather than make post hoc adjustments to the MAAS, an existing measure whose low inter-item correlation is a potential culprit for lowered fit indices in both studies. While our use of a planned missingness design yielded the same results for our hypotheses as analyses conducted using only self-report data, there were issues with the objective measures of mindfulness. The low correlation of the TMT B-A with the MAAS and resulting low factor loading on mindfulness resulted in its exclusion from analyses. The mixed results regarding objective measures and self-report measures of mindfulness is not uncommon in the literature (e.g., Anderson et al., 2007) and points to a continuing need for refinement of our understanding and measurement of mindfulness.
Third, the undergraduate sample (aged 18–20 years) used in the present study is both a strength and a limitation. This sample is drawn from among the first generational cohorts to have availability of smartphones since early adolescence and is still in the age range that shows ongoing development of the brain regions implicated in attentional control (Huttenlocker & Dabholkar, 1997) and, thus, is more vulnerable to the salient, instant, and always available rewards of smartphone use (LaRose et al., 2003). However, this use of an undergraduate convenience sample largely comprised of White, highly educated participants from a post-industrialized country may limit generalizability of the results. Nonetheless, these results are useful as they provide evidence of the relation between smartphone use and mindfulness in an at-risk age group.
Lastly, there are some concerns in the literature regarding the extent to which the MAAS captures the construct of mindfulness as opposed to mindful attention, as it fails to directly measure acceptance and non-judgmental awareness, or may simply be measuring perceived inattention (Grossman & Van Dam, 2011). Although we acknowledge the limitation of the self-reported measure, we believe the addition of our objective measures help to address some of these concerns as Brown et al. (2007) argue that there is some degree of acceptance and openness “embedded within the capacity to sustain attention to and awareness of what is occurring” (p. 245). Therefore, directing and sustaining one’s attention is indicative that one is demonstrating some degree of receptivity to the stimulus in question, and the behavioral assessments used in the current study provide a more objective measurement of these components of mindfulness.
Despite these caveats, these results offer preliminary support for our theoretical model. While we conceived of the level of involvement as increasing or decreasing the effect of smartphone use on mindfulness, we instead found evidence for the plausible alternative of involvement as a mediator of this relation. There are some fundamental tenets of this model that have not been previously studied, to our knowledge, and were not directly addressed herein. The first is that trait mindfulness is not only malleable but that it can be decreased. While there is ample support for the increase of trait mindfulness, measured both subjectively (Johnson et al., 2020; Lin et al., 2019) and objectively (e.g., Chambers et al., 2008), we found no extant studies demonstrating decreases in the capacity for mindfulness. The second is that young adults may be especially vulnerable to this potential detrimental effect of smartphone use on mindfulness, although the evidence does suggest that this capacity continues to throughout young adulthood and they have been shown to be less capable of self-regulating behavior in general (LaRose et al., 2003; Lau, 2017).
Future studies, in order to address these gaps, will require prospective designs that begin before participants have acquired smartphones. Studies that simply follow adolescents and track the development of mindfulness into early adulthood would be a welcome addition to the mindfulness literature. In order to test the suppositions of this theory, this development would need to be measured both objectively and subjectively along with objective measures of smartphone use and smartphone involvement. For comparison, it would be best to include among the cohort young adults who do not use smartphones or at least do not acquire them until late adolescence.
The results indicate a need for further research of the study variables, their relation to well-being, and subsequent implications for prevention and clinical recommendations. We discussed herein some of the extensive evidence for the link between higher levels of mindfulness and greater well-being. Future studies of the potential impact of smartphone use on mindfulness should examine how the study variables relate to well-being. The current findings can also potentially inform future studies testing the negative effects of social media use, distraction, and multitasking on productivity and well-being in educational environments (Dontre, 2021; Lau, 2017) and in the workplace (Priyadarshini et al., 2020). Furthermore, similar mechanisms as those proposed and tested in this study may parallel the literature demonstrating deleterious effects of addictive video gaming on attention and mental health (Andreassen et al., 2016), with the greatest likelihood of application for role-play games (RPGs) where social interaction is mediated through engagement in the game (Son et al., 2013). Lastly, mindfulness as a protective factor against smartphone involvement, and against potential impacts of such on well-being, should also be explored.
In conclusion, the present studies provide initial evidence that individual differences in cognitively and behaviorally involved smartphone use are associated with lower level of trait mindfulness among young adults. The study also suggests that this cognitive and behavioral involvement may be the mechanism by which smartphone use can impact mindfulness. These novel findings suggest that the constructs of smartphone use, smartphone involvement, and mindfulness deserve further attention from researchers interested in understanding the development of mindfulness and how modern technology impacts that development.
Declarations
Conflict of Interest
The authors declare no competing interests.
Ethics Approval
The Institutional Review Board at the University of South Carolina approved this study, and all ethical standards and procedures were followed throughout the study.
Informed Consent
We obtained electronic informed consent from participants for study 1 prior to beginning the online survey. After participants completed the measures, we asked for their consent to be contacted to participate in Study 2.
Disclaimer
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Artificial Intelligence
There was no use of AI tools in Study 1 or Study 2.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Met BSL Psychologie Totaal blijf je als professional steeds op de hoogte van de nieuwste ontwikkelingen binnen jouw vak. Met het online abonnement heb je toegang tot een groot aantal boeken, protocollen, vaktijdschriften en e-learnings op het gebied van psychologie en psychiatrie. Zo kun je op je gemak en wanneer het jou het beste uitkomt verdiepen in jouw vakgebied.