Abstract
Cellphone has become an integral part of people’s daily life. Anxiety may arise from cellphone separation. Although negative cognitive effects of cellphone separation were reported, the mechanism of how cellphone separation anxiety influences cognition. In the present study, we examined the effects of cellphone separation on state anxiety and physiological indicators of cognitive engagement, heart rate (HR), and high frequency heart rate variability (HF-HRV). Seventy-five college-aged participants were assigned into three groups (explicit separation, implicit separation, and control groups) and performed a Stroop task. Anxiety, HR and HF-HRV at baseline and during the task were measured. Additionally, trait anxiety, working memory capacity, and daily cellphone usage were measured as covariates. The dependent variables were submitted to repeated-measure mixed-model analysis of covariance (ANCOVA). Although there were no group effects on state anxiety and HF-HRV, the implicit cellphone separation group showed the highest level of HR response and the Stroop effect on RTs. The results suggest that the activation of the mental representation of a cellphone, which occurred in the implicit separation group, facilitates cognitive engagement and enlarge the effects of inhibition on the speed of Stroop responses.
Keywords
Introcution
Cellphones play an increasing role in its users’ daily life. With the rapid development of digital technology, cellphone has changed from unifunctional communication tool to a portable smart device that encompasses a comprehensive set of software applications (Nie et al., 2020). The prevalence of cellphone usage modifies the users’ behavior and increases their emotional attachment to the device.
As cellphones become an integral part of everyday life, the constant presence of the cellphone can make it an extension of the self as a possession. Considering the amount of personal information that is input into cellphones and the various devices (i.e., watches, tablets, headphones) that share information with cellphones about the same user, a lot of aspects of the self exist within cellphones. Thus, unwarranted misplacement of the cellphone can be interpreted as a loss of self (Belk, 1988), which may lead to negative emotions. Cellphone separation anxiety, also known as nomophobia, when it becomes more pathological, is a fear or discomfort of being without one’s mobile phone or being not able to use it (Clayton et al., 2015). Younger adults are extremely vulnerable to the negative behavioral influences of cell phone separation due to the overuse of cellphones and other smart device (Tanil & Yong, 2020).
Cellphone separation anxiety and nomophobia build from the two sequential processes of fear of missing out: perception of missing out, and then a compulsive behavior to maintain these social connections (Gupta & Sharma, 2021). Those processes induce anxiety, restlessness, and irritability. Oppositional to negative affect, the multifunctional purposes of cellphones allow for more than one medium of communication that reinforce reward. The reward experienced by the user, or the ability of users to witness the other users’ behavioral rewards is what constitutes the first processes of the perception of missing out. A study reported that the level of cellphone dependence is positively correlated with the level of anxiety experienced by college students without their phones (Bianchi & Philips, 2005). Moreover, individuals who reported higher levels of cellphone separation anxiety have been shown to have elevated levels of depression and lower levels of life satisfaction, compared to those who have less nomophobia. Therefore, nomophobia is linked to trait level individual difference in negative emotions.
In addition to negative affect, cellphone separation anxiety also influences cognitive performance. For example, without the presence of cellphones, college students exhibited worse memory task performance (Tanil & Yong, 2020). Learning and attention performance are also subject to the impacts of cellphone separation anxiety (Chen & Pai, 2018; Soomro et al., 2019).
The detrimental effects of nomophobia on cognitive processes may be due to interference of cognitive control and working memory, which is similar to the negative effects of state anxiety induced by acute stressors, or high level of trait anxiety, on working memory (Vytal et al., 2016). Cardiac indices, including heart rate (HR) and high frequency heart rate variability (HF-HRV), provide noninvasive, convenient, and objective measures of the influence of mental demand on affective and cognitive processing (Friedman, 2007; Hughes et al., 2019).
Specifically, HR indicate of a net effect of both sympathetic and parasympathetic divisions of the autonomic nervous system (ANS). While HR generally decreases when people are engaged in cognitive activities, cardiac acceleration is induced by mental stressors that are linked to excessive mental workload (Hughes et al., 2019). Moreover, several theorists have posited that, among its various features, HF-HRV is a physiological marker of cognition-emotion interaction (e.g., Thayer & Lane, 2000). HF-HRV mirrors the effect of the vagus nerve on cardiac interbeat intervals. At rest, higher levels of HF-HRV (strong vagal inputs) are associated with relaxation and signal better functional outcomes (Thayer et al., 2009). In response to cognitive demands and/or induced anxiety, decreased HF-HRV, which reflects the lessening of vagal control over cardiac functioning, is commonly (but not always) considered normative and facilitates the individual’s ability to respond to environmental demands (Siegel et al., 2018). Neuroimaging studies have confirmed that HF-HRV is associated with various brain regions that have been implicated in working memory and coping with negative emotions (Thayer et al., 2012).
Current Study
Although increasing research attention has been paid to cellphone separation in last decade, the cognitive mechanisms underlying this phenomenon remain unclear. Moreover, the identification of modifiable physiological processes will facilitate the development of possible training programs to reduce the negative effects of cellphone separation anxiety on human performance in realistic work settings. Therefore, the present study aimed to examine state anxiety and cognitive engagement, as indicated by HF-HRV, that are induced by cellphone separation in a laboratory task.
The present study includes a Stroop Task that requires participants’ conflict monitoring and inhibition control. We manipulated cellphone separation by asking a group of participants to perform the cognitive task without their phones (Explicit Separation Group), a group to perform the task with a provided phone but not their own phones (Implicit Separation Group), and a control group to perform the task with their own phones present. Participants’ anxiety levels, HR, and HF-HRV were monitored at the baseline and during the task, and their behavioral performance of the Stroop task and trait level differences in working memory, cellphone usage, and anxiety were also measured. In light of the existing literature, we hypothesized that 1) cellphone separation would induce increased anxiety in cellphone separation groups, compared to controls; 2) cellphone separation groups would show blunted HF-HRV reactivity, compared to controls; and 3) cognitive performance would be worse in the cellphone separation groups than in controls.
Method
Participants
Seventy-five participants were recruited from undergraduate psychology courses at Old Dominion University. All participants were non-smokers and were free of auditory and visual problems and mental disorders. Participants were required to abstain from alcohol for at least 12 hours and from caffeine for at least six hours prior to participation. Participants received course credit for their participation. Approval for the study was obtained from the university’s Institutional Review Board. Two participants have not completed the experimental protocol due to equipment failures, and one participant was partial color blind, which were excluded from the analysis. The final sample consists of 72 participants (Mage = 19.5 years; SD = 1.55 years; 53 female).
Task and Design
A modified Stroop task was used in the study to examine cognitive processing. The stimuli of the Stroop task were the words “Red” and “Green” that were presented on a computer screen with fonts in red or green. The task required participants to press the left arrow or the down arrow keys on the keyboard in response to the red or green font of the word, respectively (rather than the meaning of the word). The task included congruent (i.e., the meaning of the word matched the color of the font) and incongruent (i.e., the meaning of the word did not match the color of the font) conditions (see Figure 1). There was a total of 200 Stroop trials, and 100 trials for each condition. Both response speed and accuracy were emphasized in the task instruction. The stimuli were presented on the screen for 2.5 seconds, during which responses would be detected. Key pressing during the 2.5-second response window will terminate the trial, and the inter trial interval was 2.5 seconds.
Mean state anxiety, heart rate (HR), and high frequency heart rate variability (HF-HRV) change scores during the Stroop task in different groups. The comparisons were adjusted for covariates. Error bars represent ±1 standard error.
All participants were randomly assigned into one of the three groups: Explicit Separation, Implicit Separation, and Control groups. The study utilized a repeated-measure mixed design. Independent variables were the manipulation of the task (within-subject factor) and group membership (between-subjects factor). Dependent variables included self-report anxiety that were measured before and after the Stroop task, HR and HF-HRV and task performance that was assessed for each condition during the Stroop task.
Materials
Participants’ anxiety was assessed by the State-Trait Inventory (STAI; Spielberger et al., 1983). STAI included two 20-item scales to measure levels of state and trait anxiety, respectively. All items in both STAI scales are rated on a 4-point scale. STAI has been shown good internal consistency and test-retest reliability (Spielberger et al., 1983).
Several self-report questionnaires were utilized in the study. The Media and Technology Usage and Attitudes Scale (MTUAS) is a self-report questionnaire that offers a broad understanding of the usage of common forms of technology. Fifty items on a 10-point scale assessed the frequency of the technology usage behaviors. 5 items on a 9-point scale assessed the quantity of people that the user interacted with through Facebook use. Eighteen items on a 5-point scale measured the attitudes that users have pertaining to technology usage. The 15 subscales of the MTUAS provides a method of measuring technology use behaviors such as social media use, checking emails, texting, calling, etc. that would prompt smartphone users to check their phone (Rosen et al., 2013b). Due to the versatility of applications within smartphones that enable the usage behaviors included in the MTUAS, higher scores on this questionnaire could indicate higher usage of the smartphone. The questions asking about Facebook use were excluded or “Facebook” was exchanged with other types of social media which included Instagram, TikTok, and Snapchat.
The NASA – TLX measured workload on 6 subscales determining how demanding workload was during the task in mental demand, physical demand, temporal demand, performance perception, effort perception, and frustration (Hart & Staveland, 1988).
In the digit span task, participants were asked to verbally repeat a sequence of digits. The number of digits would keep increasing until the participant could not correctly repeat the digit sequence. Participants were given two initial trials at each level to complete the sequence. If one of the trials was repeated incorrectly a third trial was given at the same level. If the third trial was repeated incorrectly the task ended. If the third trial was repeated correctly the participant could advance to the next level as if they had gotten the two initial trials correct.
Procedure
Once informed consent was obtained by the appropriate procedure, the health history questionnaire was given to get some background information from the participant. After the consent procedure the participant was given the questionnaires to measure trait anxiety and cellphone usage and attitude. The digit span task was then administered. Afterwards the participants were moved to another room, and ECG electrodes were attached to the participant. To estimate resting vmHRV, participants were instructed to sit still and watch a 180-s neutral film that depicted aquatic scenes, consistent with “vanilla baseline” guidelines (Jennings et al., 1992). After the baseline, state anxiety was assessed and then the participant was randomly assigned into one of the three groups. Specifically, in the explicit separation group, participants were asked to move their cellphones to another room; in the implicit separation group, a cellphone that did not belong to the participant was placed on the desk before participants began the Stroop task, and they were also asked to move their cellphones to another room;, and in the control group, participants had their phones with them during the entire experiment. The Stroop task was then delivered after the group assignment. At the end of the task, state anxiety was assessed again. In addition, NASA-TLX was administered verbally to measure mental workload of the Stroop task. At the end of the study, participants were thanked and informed about the purpose of the study.
Physiological Recording and Data Reduction
Physiological data were collected using a BIOPAC MP160 system (BIOPAC Systems Inc, Goleta, CA). Raw signals were digitized at 1,000 Hz (16-bit) and analyzed with BIOPAC AcqKnowledge software 5.0 (BIOPAC Systems Inc, Goleta, CA). Electrocardiography (ECG) was measured with disposable, pre-gelled stress-testing spot electrodes using a modified Lead II configuration (Conmed Andover Medical, Haverhill, MA). ECG R-spikes were labelled using the AcqKnowledge software and were then inspected manually by trained raters. All participants’ ECG data met the criterion for further analyses where less than 5% signals were influenced by artifacts. HF-HRV was calculated by Kubios HRV analysis software v2.0 (Biosignal Analysis and Medical Imaging Group, Kuopio, Finland), as the power of the frequency band 0.15–0.4 Hz, using fast Fourier spectral power, which is a frequency-domain metric of cardiac vagal control, respectively (Berntson, et al., 1997). The distribution of HF-HRV values was skewed and therefore normalized with a logarithm to the base 10. Respiration was measured with a respiration transducer placed at the thoracic level. Respiration data were acquired to examine gross respiratory artifacts in HF-HRV data.
Mean accuracy and response time (RT) of congruent and incongruent trials of the Stroop task were calculated, which were then averaged for each group.
Analytic Plan
To investigate the effects of cognitive task and cellphone separation on state anxiety and cognitive engagement, anxiety scores and HF-HRV were submitted to separate repeated-measure 2 (Task) × 3 (Group) mixed-model analysis of covariance (ANCOVA). In the models, digit span, trait anxiety, and cellphone usage level were controlled for the analyses. Moreover, mean accuracy and RTs were analyzed by 2 (Condition) × 3 (Group) ANCOVA to examine groups effects on cognitive performance. Main effects and interactions on dependent variables were further examined by Helmert contrasts (F tests) and planned comparisons (paired t tests). Effect sizes of factors effects in the ANCOVA and paired t-tests were estimated with partial eta-squared and Cohen’s d, respectively.
Results
Descriptive statistics are presented in Table 1. There were no group differences in gender, age, digital span, cellphone usage, and levels of trait anxiety and state anxiety before the cognitive task. Moreover, the groups did not differ in baseline HR and HF-HRV.
Descriptive Statistics of Groups.
Note: HR = Heart rate; HF-HRV = High frequency heart rate variability; RT = Response time.
State Anxiety
The mixed ANCOVA indicates a main effect of Task on state anxiety, F(1, 68) = 4.59, p = .036, η2p = 0.07, but no group effect F(1, 68) = 0.33, p = .72, nor interaction, F(2, 68) = 0.57, p = .57 (see Figure 1).
Physiological Indicators
The results did not show indicates main effects of Task or Group on mean heart rate, ps > .54. However, there was an interaction between Task and Group, F(2, 68) = 3.24, p = .046, η2p = 0.10. Planned contrasts indicated that heart rate decreased during the Stroop task only in the implicit separation group, t(23) = -3.09, p = .009, d = 0.22, but not in two other groups, ps > .77 (see Figure 1).
There were no mains effects of Task or Group on mean heart rate, ps > .37. Moreover, the ANCOVA did not show an interaction between Task and Group, F(2, 68) = 0.37, p = .69 (see Figure 1).
Behavioral Performance during Stroop Task
The analysis did not indicate main effects or interaction on mean accuracy of the Stroop task ps > .11 (see Figure 2).
Stroop effects on behavioral performance. The comparisons were adjusted for covariates. Error bars represent ±1 standard error.
On the other hand, although there was no group effect on RTs, F(1, 68) = 0.19, p = .82, RTs were prolonged in incongruent trials, F(1, 68) = 87.84, p < .001, η2p = 0.56, compared to congruent trials. Importantly, there was an interaction on RTs, F(2, 68) = 5.35, p = .007, η2p = 0.13, and the implicit separation group showed the highest level of the Stroop effect on RTs (see Figure 2).
Discussion
The present study examined the effects of cellphone separation on affective and cognitive processes. A modified Stroop task was used to induce cognitive inhibition, which requires top-down control. Cellphone separation was manipulated as explicit and implicit conditions (Tanil & Yong, 2020), and state anxiety, cardiac indices, and behavioral performance of the Stroop task were assessed. In addition, trait level anxiety and smart mobile device usage were measured. Our hypotheses were partially supported. Although there were no group effects on state anxiety and HF-HRV, the implicit cellphone separation group showed the highest level of HR response and the Stroop effect on RTs.
Specifically, the hypothesis about state anxiety was not supported. Anxiety level was increased during the Stroop task, which is consistent with prior studies (e.g., Richards et al., 1992). Moreover, while those studies utilized emotional Stroop tasks, we did not include affective information in the Stroop task. Thus, our results indicate that a “pure” cognitive process induces negative feelings and can be explained by that mental workload induced by cognitive inhibition is associated with anxiety. However, no group effect on the anxiety increase was evident, suggesting that cellphone separation did not produce additional effects on state anxiety. This null finding may be attributed to the discrepancy between cognitive processes involved in cellphone use and the Stroop task. In other words, both explicit and implicit cellphone separation did not influence inhibition. Alternatively, participants might be engaged by the Stroop task, and there were only limited mental recourses allotted for processing cellphone separation.
Our observations are consistent with the possible explanation. Several of the participants in the implicit separation group explained during the debriefing process at the end of the study that they thought the inactive, powerless phone placed on the desk they were seated at was surveilling them in some way. The participants in the sample are university students and have an awareness that there is deception involved in scientific research and seem to have assumed that the unexplained cell phone that was in the room with them was part of the deception. However, there is a pattern showing a slightly higher increase in the increase in state anxiety from the explicit and the implicit separation groups to the control group.
The hypotheses regarding cardiac indices were partially supported While HR did not significantly change in three groups overall, the interaction and planned contrasts indicated that the implicit separation group showed HR decrease during the Stroop task. This finding may be explained by that the participants in the implicit separation group were highly engaged in the Stroop task, as slowing HR is an important component of cognitive inhibition (Thayer & Lane, 2009). However, it is surprising that this effect was absent in two other groups. A fake cellphone might serve as a replacement of the mental representation of the participant’s own cellphone, and this mental process was unconscious and prevented other distraction during the task. Unlike HR, HF-HRV during the Stroop task did not change from the baseline or differ between groups. HF-HRV reflects the activation level of the parasympathetic division of the ANS, and the present results suggest that although the mental workload increased anxiety, inhibition was not a stressor and did not elicit physiological stress responses.
The results of the Stroop effect on RT for the implicit separation group is consistent with the heart rate findings. Stroop inhibition is the largest with the implicit separation group and reaction times were longer. The Stroop task was used to test the findings of Hartanto and Yang (2016). In Hartanto and Yang (2016) the condition that was separated from their phones during a Stroop task also had longer reaction times in the condition that experienced cellphone separation. The reasoning for having a condition where a device was present on the desk that participants were completing tasks at came from Hartmann et al. (2020). Although the study of Hartmann et al. (2020) measured the effects of the presence of a smartphone on a desk on prospective and short-term memory tasks rather than reaction time and accuracy tasks used by the current study. The presence of the cellular device on a desk while completing cognitive tasks had a different effect in the current study. We found that there was an effect on the data when a cell phone was present on the desk during the Stroop tasks.
The present study has several strengths, including utilizing both implicit and explicit cellphone separation manipulations, collecting physiological measures to index anxiety and task engagement, and controlling for individual differences in levels of anxiety and cellphone usage. However, our findings need to be evaluated with considering a number of limitations. For example, there was no assessment of participants’ attentional processes during the task. Future studies may include eye tracking in cognitive tasks. Moreover, the current study focused on college-aged individuals. There may be cohort effects on cellphone usage and separation anxiety. In addition, future studies could utilize other cognitive tasks to gather behavioral data. Less invasive physiological devices could be used such as a portable EEG and are a feasible alternative method to collect physiological data. To increase external validity the setting of future experiments could be in a workplace or school setting.
Conclusion
Cellphones plays an important role in everyday life, and nomophobia have both psychological and behavioral impacts on ongoing tasks. By manipulating the presence of personal and a simulated cell phone, we have examined effects of cellphone separation on state anxiety, cardiac indices of task engagement, and behavioral performance that required inhibition control. We found that heart rate was decreased for the implicit separation group coinciding with the reaction times being the highest for the same group. A pattern of higher increases in state anxiety scores were found in both experimental groups after separation from personal cellphones. The results suggest that the activation of the mental representation of a cellphone, which occurred in the implicit separation group, facilitates cognitive engagement and enlarge the effects of inhibition on the speed of Stroop responses. Future studies are expected to replicate our findings in real-life settings.