Abstract
Background and aims
Previous evidence has indicated that problematic social media use (PSMU) is characterized by an attentional bias to social media icons (such as Facebook icons), but not to social webpages (such as Facebook webpages). They suggest that there may be other factors influencing attentional bias like fear of missing out (FoMO). But it remains unclear how FoMO moderates attentional bias in PSMU. This study aims to investigate whether PSMU show attentional bias for stimuli associated with social media, and how FoMO moderates on attentional bias among PSMU through experimental methods.
Methods
Based on the Interaction of Person-Affect-Cognition-Execution (I-PACE) model, this study explored mechanisms of attentional bias to social media icons (such as WeChat) related to PSMU and further examined the role of FoMO in this relationship. Specifically, attentional bias patterns to social media icons of 62 participants (31 PSMU and 31 control group) were explored during a dot-probe paradigm combined with eye-tracking in Experiment 1, and attentional bias patterns to social media icons of another 61 individuals with PSMU with different FoMO levels was explored during a dot-probe paradigm combined with eye-tracking in Experiment 2.
Results
Results revealed that individuals with PSMU had an attentional bias toward social media icons, demonstrated by attentional maintenance, and such bias such bias was moderated by FoMO negatively, demonstrated by attentional vigilance and maintenance in PSMU/high FoMO.
Conclusion
These results suggest that attentional bias is a common mechanism associated with PSMU, and FoMO is a key factor on the development of PSMU.
Introduction
About 5 billion people use a social media service worldwide as of January 2024, with China accounting for the largest portion of global social media use (Statista, 2024). Though social media use can enhance social connections (Fumagalli, Dolmatzian, & Shrum, 2021) and provide opportunities for self-presentation (Wang et al., 2018), it can also increase the likelihood of behavioral dysfunction (Soh, Charlton, & Chew, 2014) and can result in problematic social media use (Kuss & Griffiths, 2017). Problematic social media use (PSMU) refers to an uncontrolled and insatiable desire to use social media (Miranda, Trigo, Rodrigues, & Duarte, 2023), which is thought to be similar to a behavioral addiction (Bielefeld et al., 2017). It has become a relatively common social phenomenon because of the widespread use of mobile networks (Jiang, Bai, Bai, Qi, & Wu, 2019) and proliferation of the smartphone to access social media (Sha, Sariyska, Riedl, Lachmann, & Montag, 2019). However, neither the International Classification of Diseases (ICD-11, (WHO, 2023) nor Diagnostic and Statistical Manual of Mental Disorders (DSM-5, APA, 2013) include PSMU as a distinct diagnosis, instead suggesting that it could be categorized as a subtype of Internet addiction for further research (Brand et al., 2022). For a taxonomy of Internet Use Disorders see Montag, Wegmann, Sariyska, Demetrovics, and Brand (2021).
Previous studies have found that attentional bias toward addictive cues is one of the specific characteristics of addictive disorders (Field et al., 2016), thought to play a major role in the development of addiction (Jones, Worrall, Rudin, Duckworth, & Christiansen, 2021). In addiction, attentional bias refers to the process of preferentially seeking and allocating more attention when addiction-related cues are presented (Field & Cox, 2008), which contains three components, attentional vigilance (a process of prioritizing attention to specific stimuli, which facilitates the detection of specific information), maintenance (a process of stopping attentional processing of a specific stimulus difficultly), and avoidance (a process of allocating attention by preferentially avoiding specific stimuli and directing attention to non-specific stimuli) (Sheppes, Luria, Fukuda, & Gross, 2013). One of the key factors to assess addiction clinically is a persistent and stable attentional bias toward addiction-related stimuli (Anderson, 2016). So, it is important to support the assessment of PSMU by exploring whether stable attentional bias towards social media icons is present (Brand et al., 2022). Interestingly, the results of attentional bias involved with PSMU are not consistent. Jiang, Bai, et al. (2019) found a relationship between PSMU and attentional bias towards social text stimuli through a spatial cuing task and response time calculations. Thomson, Hunter, Butler, and Robertson (2021) did not find any evidence for a relationship between social media use and attentional bias to social media stimuli through an attentional capture task and response time calculations. Nikolaidou and colleagues found that PSMU only showed an association with attentional bias towards social icon stimuli through a dot-probe task and eye movement indicators calculated (Nikolaidou, Fraser, & Hinvest, 2019). It can be seen that the three studies conducted different paradigms, so it is unclear whether the incongruency of findings was caused by the different paradigms. The dot-probe task can yield different components of attentional bias through computation (MacLeod & Mathews, 1988). Further, eye-tracking may reflect attentional bias in a dynamical way through different indicators, which may compensate for the static response process of attentional bias by single reaction time (Jiang, Wang, & Qiang, 2019). Also, previous studies found that problematic internet users (hence the umbrella term comprising social media) may allocate more attention to internet-related stimuli, leading to difficulty in shifting attention away from related cues (Field & Cox, 2008). So, this study explored attentional bias related to PSMU by a dot-probe task combined with eye-tracking to learn about the patterns of attentional bias among individuals with PSMU towards social media icons.
The Interaction of Person-Affect-Cognition-Execution (I-PACE) model (Brand et al., 2016, 2019) indicates that individuals develop specific cognitive biases related to social media use (see also, Davis, 2001), further influencing the “first choice” of specific online behaviors and leading to a stable attentional bias towards social media (Brand et al., 2019). Also, incentive sensitization theory proposes that the repeated behavior of use may produces incremental neuroadaptations in reward circuitry, rendering it increasingly and perhaps permanently, hypersensitive (‘sensitized’) and lower top-down controls to associated stimuli (Robinson & Berridge, 2003). Further, sensitization of the incentive salience (for ‘wanting’) can occur independently of changes in neural systems that mediate the subjective pleasurable effects (‘liking’) and of neural systems that mediate withdrawal (Robinson & Berridge, 2003). Thus, sensitization of the incentive salience can produce addictive behavior (compulsive seeking and taking), which may lead to difficulty in shifting attention to related cues and developing into attention maintenance (see also, Field & Cox, 2008). Studies on PSMU have found that there was impaired top-down controls of PSMU (Nasser et al., 2020; Su et al., 2021), and based on this, this study proposed hypothesis 1, that individuals with PSMU would have attentional maintenance towards social media icons.
The use of social media will allow a specific cognitive-behavioral pattern (like attentional bias toward relevant stimuli), but a specific anxiety may also be caused by it because of too much information (Bandura, 2001), such as the fear of missing out (Przybylski, Murayama, DeHaan, & Gladwell, 2013). Fear of missing out (FoMO) is a specific anxiety in the context of social media use (Altuwairiqi, Jiang, & Ali, 2019), referring to “a pervasive apprehension that others might be having rewarding experiences from which one is absent”, and a desire to stay connected continually with others (Przybylski et al., 2013, p. 1841). Studies have found that FoMO was a significant predictor of PSMU (Opsenica Kostić, Pedović, & Stošić, 2022), and FoMO was a moderator in the development of PSMU (Wang et al., 2018). However, it may not fully explain the mechanisms of how FoMO facilitates the development of PSMU (Elhai et al., 2021). In terms of definition, FoMO is a type of anxiety caused by inability to access information about others (Przybylski et al., 2013). High levels of anxiety may impact attentional allocation and show an attentional bias toward relevant stimuli (Eysenck, Derakshan, Santos, & Calvo, 2007). Similar to anxiety, high levels of FoMO showed an abnormality in structure or function of brain networks related to attentional vigilance (Yin et al., 2023) and inhibitory control (Xu, Chen, & Tian, 2024). And high FoMO affects attentional allocation under a difficult task situation, producing maintenance from social reward stimuli (Matias et al., 2021). Although previous work found that the level of FoMO may influence performance of attentional bias among individuals with PSMU (Hou, Zhu, & Fang, 2021), there is no direct evidence on how FoMO moderated attentional bias in PSMU.
According to the I-PACE model, FoMO can play a moderating role on cognitive responses (e.g., attentional bias) of PSMU (Brand et al., 2019). High levels of FoMO may meet psychological needs by using social media (Brand et al., 2019), and further may enhance an urge of reward-seeking and reduction of inhibitory control (Robinson & Berridge, 2003), leading to a stable attentional bias towards social media icons among PSMU (Field & Cox, 2008). Based on the I-PACE model and incentive-sensitization theory, this study proposed hypothesis 2, that FoMO plays a moderating role in the attentional bias of PSMU, whereby high FoMO individuals with PSMU show greater attentional vigilance and maintenance towards social media icons compared to the low FoMO group.
Considering this, two experiments were conducted to explore how FoMO moderated the attentional bias among individuals with PSMU. Experiment 1 investigated the attentional bias toward social media icons among those with PSMU through a visual dot-probe task. Experiment 2 investigated how different levels of FoMO affect attentional bias toward social media icons among PSMU. Furthermore, as Nikolaidou et al. found that social media icons may be more appropriate to investigate the attentional bias among those with PSMU (Nikolaidou et al., 2019), this study used social media icons as experimental stimuli.
EXPERIMENT 1
Methods
Participants
The sample size was set based on two aspects: (1) An a priori power analysis was conducted using G*Power 3.1 (f at 0.25, α at 0.05, power at 0.80), and it revealed that a minimum total sample of 34 participants was required in a 2 × 2 mixed design. (2) Previous studies on attentional bias among individuals with PSMU resulting in a significant different had sample sizes ranging from 20 to 30 participants in each group (Jiang, Bai, et al., 2019; Nikolaidou et al., 2019; Zhao et al., 2022) (same with Experiment 2).
There were no missing data because the Web questionnaire was set up to not allow skipping of any items. After excluding 3 participants who chose “do not agree to participate the further experiment” on the first web page, a total of 198 participants provided informed consent and were screened for PSMU in Experiment 1. Thirty-one participants were classified in the PSMU group because of Bergen Social Media Addiction Scale (BSMAS) scores of 3 or above on (at least choose “Sometimes”, “Often” or “Very often”) at least four of the six items (Andreassen, Torsheim, Brunborg, & Pallesen, 2012), and thirty-one participants were classified as the control group because BSMAS scores were not 3 or above on (choose “Very rarely” or “Rarely”) at least four of the six items (Andreassen et al., 2012). So, 62 participants were recruited at a university in Tianjin, China, 8 males (12.9%) and 54 females (87.1%), aged between 18 and 21 years (Mage = 18.68, SD = 0.62). Table 1 provides a detailed description of the sample in Experiment 1. All participants were right-handed, with no mental or physical disorder, with normal vision or corrected vision.
Demographic information in Experiment 1
PSMU | Control | t/χ2 | p | Post-hoc comparisons | |||
Age | N | 31 | 31 | 1.232 | 0.223 | ||
Mean | 18.77 | 18.58 | |||||
SD | 0.72 | 0.50 | |||||
Min | 18 | 18 | |||||
Max | 21 | 19 | |||||
Gender | Male | N | 3 | 5 | 0.574 | 0.707 | |
Total N | 31 | 31 | |||||
Proportion | 0.10 | 0.16 | |||||
Female | N | 28 | 26 | ||||
Total N | 31 | 31 | |||||
Proportion | 0.90 | 0.84 | |||||
BSMAS | N | 31 | 31 | 13.834 | <0.001 | PSMU > Control | |
Mean | 24.19 | 13.84 | |||||
SD | 3.04 | 2.85 | |||||
Min | 21 | 7 | |||||
Max | 30 | 19 |
Note: PSMU: problematic social media users; SD: standard deviation; BSMAS: the Bergen Social Media Addiction Scale; *p < 0.05; **p < 0.01; ***p < 0.001; Age and BSMAS were test with independent t-tests, gender was test with χ2 tests.
Measures
Bergen Social Media Addiction Scale (BSMAS; Andreassen et al., 2012; Luo et al., 2021). A Chinese version of the BSMAS (Luo et al., 2021) assesses the severity of PSMU. There are six items (e.g., “Spent a lot of time thinking about or planned use of social media?”) using a five-point Likert-type scale from 1 (very rarely) to 5 (very often). Total scores range from 6 to 30. The PSMU group was identified as scoring 3 or above on (at least choose “Sometimes”, “Often” or “Very often”) at least four of the six items (Andreassen et al., 2012). The initial internal consistency coefficient of this scale was α = 0.83 (Luo et al., 2021). In this study, Cronbach's α was 0.853, and McDonald's ω was 0.855.
Materials
Experimental stimuli consisted of social app icons (WeChat, Weibo, etc.), and non-social app icons (Meitu, 12306, etc.), which were retrieved from a Baidu web search and reprocessed by Photoshop CC (2019) in the size of 105 × 105 mm. Virtual icons as neutral stimuli were processed by Photoshop CC (2019) in the size of 105 × 105 mm, too. Virtual icons were created with color and complexity, considered to pair with social and non-social icons. 66 college students who did not participate in the study were recruited to rate the stimuli in terms of representation (1 = low represent of social app, 5 = high represent of social app), familiarity (1 = very unfamiliar; 5 = very familiar), and arousal (1 = very low, 5 = very high). Finally, 20 social app icons and 20 non-social app icons were selected, and matched 40 virtual icons as neutral stimuli. The results of selected stimuli ratings were tested using SPSS (version 26.0, IBM, USA) and showed that there was a significant difference in representation between social media icons and non-social icons (t(38) = 2.378, p = 0.009, Cohen's d = 0.85). No significant difference was found in familiarity (t(38) = 0.891, p = 0.378, Cohen's d = 0.29) or arousal (t(38) = 1.032, p = 0.309, Cohen's d = 0.34) between social media icons and non-social icons. And no significant difference was found for matched neutral icons in representation (t(38) = 1.118, p = 0.271, Cohen's d = 0.35), familiarity (t(38) = 1.032, p = 0.309, Cohen's d = 0.33) or arousal (t(38) = 0.148, p = 0.883, Cohen's d = 0.05). The statistical results are displayed in Suppl. Table 1.
Procedure
Participants were sampled from a large university in Tianjin, China. Students18 years or older were eligible to participate this study. Data collection involved two stages: (1) The participants answered questions concerning age and gender, and the BSMAS (Andreassen et al., 2012; Luo et al., 2021) through a link to the Web questionnaire (www.wjx.cn). The information statement describing the study aims was provided on the first page of the questionnaires, and participants provided informed consent to participate by clicking a button online. (2) Participants who met criteria for the PSMU and Control groups were contacted to participate in an eye-tracking experiment via a pre-determined web contact to explore attentional bias. Specifically, a Visual Dot-Probe Task (MacLeod, Mathews, & Tata, 1986) was performed on Eye-link 1000 plus, which was developed by SR Research, Canada, with a sampling rate of 1000 Hz. Prior to the experiment, participants familiarized themselves with the task in a practice session using 4 trials that were not used in the formal experiment. For the experiment, instructions of this study were presented on the screen. A five-point calibration was performed after participants understood the procedures. A fixation cross was first presented in the center of the screen, and participants were asked to focus on the center of it. Then, stimuli trials with 20 social app icons, 20 non-social app icons and matched 40 virtual icons respectively were presented. Each trial lasting for 2000 ms was presented in a pair of social-neutral icons or non-social-neutral icons as stimuli; a probe dot appeared following either on the left or right side of the computer screen replacing one of the icons, resulting in 160 stimuli overall. After the stimuli trial disappeared, a probe dot appeared, and participants were asked to press the “F” key if the dot was presented on the left, or to press the “J” key if the dot was presented on the right. Finally, a blank screen was presented lasting for 1000 ms. The stimuli were presented on a 24-inch display with a resolution of 1920 × 1280 pixels. The participants sat comfortably in a chair with the head placed 60 cm from the screen, and a desktop infrared eye-tracking was used to track the eye movements of participants when they were looking at screen. Stimuli were presented in a pseudorandom way, and the whole experiment lasted about 15 min (Fig. 1).
Statistical analysis
SPSS was used for analysis. Age and scores of the BSMAS scale were analyzed with independent t-tests, while χ2 tests were conducted to compare the association between PSMU groups and gender.
We excluded trials with reaction times (RT) lower than 200 ms or higher than 2000 ms, as well as remaining trials below or above 3 standard deviations from each participant's means (Pabst et al., 2023). We calculated scores of attentional bias by referring to the formula of MacLeod and Mathews (1988) to identify the components of attentional bias. Scores of attentional bias = [(LP/RT-LP/LT) + (RP/LT-RP/RT)]/2, whereby L means left, R means right, P means the probe dot, and T means the target stimuli like social media icons or non-social icons (MacLeod & Mathews, 1988). Then, repeated measures of ANOVA were used to calculate the different of scores of attentional bias. And, a one-sample t-test was conducted on the scores of attentional bias (with the threshold set at 0); it represented attentional maintenance if the scores of attentional bias was greater than 0 significantly, it represented attentional avoidance if the scores of attentional bias was lower than 0 significantly, and it represented no attentional bias if there was no different between the scores of attention bias and 0 (MacLeod & Mathews, 1988).
We excluded eye-tracking indicators lasting less than 100 ms or more than 1000 ms, as well as remaining trials below or above 3 standard deviations from each participant's means (Gao, Li, Bai, & Gao, 2023). And we calculated scores of attentional bias of the first fixation latency, first fixation duration, and dwell time by referring to the formula of Kou (Kou, Su, Luo, & Chen, 2015) to identify the components of attentional bias. Scores of attentional bias of the first fixation latency = the first fixation latency of target stimuli (like social media icons or non-social icons) - the first fixation latency of neutral stimuli (Kou et al., 2015). Scores of attentional bias of the first fixation latency are reactive to the early attentional orienting acceleration to the stimuli which reflect attentional vigilance effects (Jiang, Wang, & Qiang, 2019); it represented attentional orienting acceleration if the score was lower than 0 significantly, it represented attentional orienting deceleration if the score was greater than 0 significantly, and it represented no attentional bias if there was no different between the score and 0 (Kou et al., 2015). Scores of attentional bias of the first fixation duration = the first fixation duration of target stimuli (like social media icons or non-social icons) - the first fixation duration of neutral stimuli (Kou et al., 2015). Scores of attentional bias of the first fixation duration are reactive to the early attentional process to the stimuli which reflect attentional avoidance effects (Jiang, Wang, & Qiang, 2019); it represented early attentional avoidance if the score was lower than 0 significantly, it represented early attentional maintenance if the score was greater than 0 significantly, and it represented no attentional bias if there was no different between the score and 0 (Kou et al., 2015). Scores of attentional bias of the dwell time = the dwell time of target stimuli (like social media icons or non-social icons)/(the dwell time of target stimuli + the dwell time of neutral stimuli) (Kou et al., 2015). Scores of attentional bias of the dwell time are reactive to the late attentional process to the stimuli which reflect attentional disengagement effects (Jiang, Wang, & Qiang, 2019); it represented attentional maintenance (or difficulty in attention disengagement) if the score was greater than 50.0% significantly, it represented attentional avoidance if the score was lower than 50.0% significantly, and it represented no attentional bias if there was no different between the score and 50.0% (Kou et al., 2015). Then, repeated measures ANOVAs were used to calculate the difference of scores of attentional bias of the first fixation latency, first fixation duration, and dwell time. And one-sample t-test was conducted on the scores of attentional bias (with the threshold set at 0 or 50.0%, respectively.
Finally, correlational analysis and Fisher r-z were used to explore the relationship between two groups on attentional bias patterns toward the same stimuli.
Ethics
The study procedures were carried out in accordance with the Declaration of Helsinki. The Ethics Committee of Tianjin Normal University approved the study. All participants were informed about the study, and all provided informed consent.
Results
After excluding trials with the criteria of invalid data, the reaction accuracy of the valid data in this study was 100%, and only the reaction time data were analyzed. And, excluding trials below or above 3 standard deviations from each participant's means (Pabst et al., 2023), 97.4% of trials were analyzed.
Reaction time
A 2 (group: problematic/control) × 2 (type of stimuli: social/non-social) repeated measures ANOVA was used to examine the scores of attentional bias of RTs (The RTs and statistical results are displayed in Suppl. Table 2, Suppl. Table 3 and Suppl. Table 4).
Scores of attentional bias of RT. Repeated measures ANOVA for the scores of attentional bias of first fixation latency showed an interaction effect between group and type of stimuli, F(1, 60) = 19.361, p < 0.001, ηp2 = 0.244. Simple effects analyses found that under conditions of social icons, PSMU group had higher scores of attentional bias compared to the control group, F(1,60) = 24.810, p < 0.001, η2 = 0.293, under conditions of non-social icons, PSMU group also had higher scores of attentional bias compared to the control group, F(1,60) = 4.548, p = 0.037, η2 = 0.070; in PSMU group, scores of attentional bias towards social media icons was higher than non-social icons, F(1,60) = 12.066, p = 0.001, η2 = 0.167, in control group, scores of attentional bias towards social media icons was lower than non-social icons, F(1,60) = 7.558, p = 0.008, η2 = 0.112. But the main effect of group and the main effect of stimuli were non-significant, F(1,60) = 4.001, p = 0.050, η2 = 0.063, F(1, 60) = 0.262, p > 0.05, ηp2 = 0.004, respectively.
Direct comparisons between the scores of attentional bias and 0 were performed in each group. In the PSMU group, the scores of attentional bias were significantly higher than 0 only for social media icons, t(30) = 5.319, p < 0.001, Cohen's d = 0.95, but not for non-social icons, t(30) = −0.684, p > 0.05, Cohen's d = 0.12. In the control group, the scores of attentional bias were significantly higher than 0 only for non-social icons, t(30) = 2.876, p = 0.007, Cohen's d = 0.52, but not for social media icons, t(30) = −1.996, p > 0.05, Cohen's d = 0.36. These results indicated that the PSMU group had attentional maintenance to social media icons, and the control group had attentional maintenance to non-social icons.
The eye-tracking indicators
A 2 (group: problematic/control) × 2 (type of stimuli: social/non-social) repeated measures ANOVA was used to examine scores of attentional bias of the eye-tracking indicators (The statistical results are displayed n Suppl. Table 5, Suppl. Table 6 and Suppl. Table 7).
Scores of attentional bias of first fixation latency
Repeated measures ANOVA for the scores of attentional bias of first fixation latency showed a main effect for type of stimuli, F(1, 60) = 5.727, p = 0.020, ηp2 = 0.087, indicating that the scores of attentional bias of first fixation latency was greater for social media icons. While the main effect of group, F(1, 60) = 0.016, p > 0.05, ηp2 = 0.000, and the interaction between group and type of stimuli, F(1, 60) = 1.054, p > 0.05, ηp2 = 0.017 were non-significant.
Scores of attentional bias of first fixation duration
Repeated measures ANOVA for the scores of attentional bias of first fixation duration showed a main effect for type of stimuli, F(1, 60) = 6.417, p = 0.014, ηp2 = 0.097, indicating that the scores of attentional bias of first fixation duration were higher for non-social icons. While the main effect of group, F(1, 60) = 1.014, p > 0.05, ηp2 = 0.017, and the interaction between group and type of stimuli, F(1, 60) = 0.298, p > 0.05, ηp2 = 0.005 were non-significant.
Scores of attentional bias of dwell time
Repeated measures ANOVA for the scores of attentional bias of dwell time showed an interaction effect between group and type of stimuli, F(1, 60) = 8.162, p = 0.006, ηp2 = 0.120. Simple effects analyses found that under conditions of social icons, PSMU group had higher scores of attentional bias compared to the control group, F(1,60) = 4.495, p = 0.038, η2 = 0.070; also, in control group, scores of attentional bias towards social media icons was lower than non-social icons, F(1,60) = 11.208, p = 0.001, η2 = 0.157. While the main effect of group, F(1, 60) = 1.070, p > 0.05, ηp2 = 0.018, and the main effect of type of stimuli, F(1, 60) = 3.526, p > 0.05, ηp2 = 0.056 were non-significant.
Direct comparisons between scores of attentional bias and 50.0% were performed in each group. In the PSMU group, scores of attentional bias were significantly higher than 50.0% only for social media icons, t(30) = 2.324, p = 0.027, Cohen's d = 0.33, but not for non-social icons, t(30) = 1.059, p > 0.05, Cohen's d = 0.19. In the control group, scores of attentional bias were not significantly higher than 50.0% for social media icons, t(30) = −1.564, p > 0.05, Cohen's d = 0.28, and non-social icons, t(30) = 1.667, p > 0.05, Cohen's d = 0.30. These results indicated that the PSMU group had an attentional maintenance to social media icons (Fig. 2).
Correlation analyses
The correlational results indicated that correlation trends between BSMAS scores and scores of attentional bias were different for different groups, supporting the idea that there might be different attentional bias patterns for different groups when they were faced with the same stimuli. Specifically, correlation analysis between BSMAS scores and the scores of attentional bias of dwell time in social media icons showed a positively relationship in the PSMU group (r = 0.517, p = 0.003), but no significant relationship in control group (r = 0.155, p = 0.406), with Fisher's Z test finding no significant difference between two correlation coefficients (p = 0.059). Correlation analysis between BSMAS scores and the scores of attentional bias of dwell time in non-social icons showed a positive relationship in the PSMU group (r = 0.490, p = 0.005), but no significant relationship in control group (r = 0.145, p = 0.436). Fishers Z test found no significant difference between two correlation coefficients (p = 0.072). Correlation analysis between BSMAS scores and the scores of attentional bias of RT in non-social icons showed a negative relationship in the control group (r = −0.376, p = 0.037), but no significant relationship in the PSMU group (r = −0.098, p = 0.600). Fisher's Z test found no significant difference between two correlation coefficients (p = 0.133) (Fig. 3).
Discussion for Experiment 1
Experiment 1 found that not only RT but also dwell time indicated that individuals with PSMU showed an attentional maintenance toward social media icons, which supported hypothesis 1. This finding suggests that PSMU may involve a difficulty in shifting attention to social media icons because of a reduction in inhibitory control (Robinson & Berridge, 2003), developing a stable attentional bias toward social media icons over the course of long-term social media use (Brand et al., 2016, 2019). Also, the correlational results indicated that correlation trends between BSMAS scores and scores of attentional bias were different between PSMU and control groups, supporting that the attentional bias patterns of PSMU may be specific to a certain extent.
EXPERIMENT 2
Methods
Participants
The sample size was set with the same basis of Experiment 1.
There were no missing data because the Web questionnaire was set up to not allow skipping of any items. After excluding 6 participants who chose “do not agree to participate the further experiment” at the first web page, a total of 251 people provided informed consent and were screened in Experiment 2. Twenty-nine participants were classified as the PSMU/high FoMO group (BSMAS scores of 3 or above on at least four of the six items which means at least choose “Sometimes”, “Often” or “Very often”, Andreassen et al., 2012; and total scores in the top quartile on the FoMO scale), and thirty-two participants were classified as the PSMU/low FoMO group (BSMAS scores of 3 or above on at least four of the six items which means at least choose “Sometimes”, “Often” or “Very often”, Andreassen et al., 2012; and total scores in the lower quartile on the FoMO scale). So, 61 participants were recruited at a university in Tianjin, China, with 10 males (16.4%) and 51 females (83.6%), aged between 18 and 23 years (Mage = 19.54, SD = 1.34). Table 2 provide detailed description of the sample in Experiment 2. All participants were right-handed, with no mental or physical disorder, and normal vision or corrected vision.
Demographic information in Experiment 2
PSMU/high FoMO | PSMU/low FoMO | t/χ2 | p | Post-hoc comparisons | |||
Age | N | 29 | 32 | −0.468 | 0.223 | ||
Mean | 19.34 | 19.72 | |||||
SD | 1.40 | 1.28 | |||||
Min | 18 | 18 | |||||
Max | 23 | 22 | |||||
Gender | Male | N | 4 | 6 | 0.273 | 0.735 | |
Total N | 29 | 32 | |||||
Proportion | 0.14 | 0.19 | |||||
Female | N | 25 | 26 | ||||
Total N | 29 | 32 | |||||
Proportion | 0.86 | 0.81 | |||||
BSMAS | N | 29 | 32 | 1.594 | 0.060 | ||
Mean | 23.00 | 21.41 | |||||
SD | 3.86 | 2.56 | |||||
Min | 19 | 18 | |||||
Max | 30 | 28 | |||||
FoMO | N | 29 | 32 | 12.893 | <0.001 | PSMU/high FoMO > PSMU/low FoMO | |
Mean | 42.86 | 29.97 | |||||
SD | 3.30 | 2.96 | |||||
Min | 40 | 22 | |||||
Max | 50 | 34 |
Note: PSMU/high FoMO: high FoMO problematic social media users; PSMU/low FoMO: low FoMO problematic social media users; SD: standard deviation; BSMAS: the Bergen Social Media Addiction Scale; FoMO: the Fear of missing out Scale; *p < 0.05; **p < 0.01; ***p < 0.001; Age, BSMAS and FoMO were tested with independent t-tests, gender was tested with χ2 tests.
Measures
Bergen Social Media Addiction Scale (BSMAS; Andreassen et al., 2012; Luo et al., 2021). A Chinese version of the BSMAS (Luo et al., 2021) assessing the severity of PSMU was the same as in Experiment 1. Also, t here are six items (e.g., “Spent a lot of time thinking about or planned use of social media?”) using a five-point Likert-type scale from 1 (very rarely) to 5 (very often). Total scores range from 6 to 30. The PSMU group was identified as scoring 3 or above on (at least choose “Sometimes”, “Often” or “Very often”) at least four of the six items (Andreassen et al., 2012). In this study, Cronbach's α of BSMAS was 0.833, and McDonald's ω was 0.835.
Fear of Missing Out Scale (FoMO scale; Wang et al., 2022). A Chinese version of the FoMO scale was used to investigate FoMO levels of participants (Wang et al., 2022). There are ten items (e.g., “I fear others have more rewarding experiences than me”) using a five-point Likert-type scale from 1 (not at all true of me) to 5 (extremely true of me). Total scores range from 10 to 50. The higher the score, the higher the FoMO level. The initial internal consistency coefficient of this scale was α = 0.84 (Wang et al., 2022). In this study, Cronbach's α was 0.883, and McDonald's ω was 0.882.
Materials
Experimental stimuli in Experiment 2 were same as Experiment 1.
Procedure
Experiment 2 followed the same procedure as Experiment 1, with a notable exception. Specifically, participants completed a Chinese version of the FoMO scale (Przybylski et al., 2013; Wang et al., 2022) in Experiment 2.
Statistical analysis
Statistical analysis and the exclusion criteria of invalid data were same as Experiment 1, with a notable exception. Specifically, age, and scores on the FoMO scale were analyzed with independent t-tests, while χ2 tests were conducted to compare the association between FoMO group and gender.
Ethics
The ethics statement is the same as in Experiment 1.
Results
After excluding trials with the criteria of invalid data, the reaction accuracy of the valid data in this study was 100%, and only the reaction time data were analyzed. And, after excluding trials below or above 3 standard deviations from each participant's means (Pabst et al., 2023), 95.2% of trials were analyzed.
Reaction time
A 2 (group: PSMU/high FoMO vs. PSMU/low FoMO) × 2 (type of stimuli: social vs. non-social) repeated measures ANOVA was used to examine the scores of attentional bias of RTs (The RTs and statistical results are displayed in Suppl. Table 8 and Suppl. Table 9).
The was no interaction effect between group and type of stimuli, F(1, 59) = 1.132, p > 0.05, ηp2 = 0.019, no main effect of group, F(1, 59) = 0.231, p > 0.05, ηp2 = 0.004, and no main effect of type of stimuli, F(1, 59) = 0.007, p > 0.05, ηp2 = 0.001.
The eye-tracking indicators
A 2 (group: PSMU/high FoMO vs. PSMU/low FoMO) × 2 (type of stimuli: social vs. non-social) repeated measures ANOVA was used to examine scores of attentional bias of the eye-tracking indicators (The statistical results are displayed in Suppl. Table 10, Suppl. Table 11 and Suppl. Table 12).
Scores of attentional bias of first fixation latency. Repeated measures ANOVA for the scores of attentional bias of first fixation latency showed a significant main effect of group, F(1, 59) = 5.042, p = 0.029, ηp2 = 0.079, indicating that the scores of attentional bias of first fixation latency were lower for the PSMU/high FoMO group. The interaction between group and type of stimuli was significant, F(1, 59) = 10.281, p = 0.002, ηp2 = 0.148. Simple effects analyses found that under conditions of social icons, the PSMU/high FoMO group had lower scores of attentional bias compared to the PSMU/low FoMO group, F(1,59) = 14.579, p < 0.001, η2 = 0.198; in the PSMU/high FoMO group, scores of attentional bias towards social media icons was lower than non-social icons, F(1,59) = 5.950, p = 0.018, η2 = 0.092, in the PSMU/low FoMO group, scores of attentional bias towards social media icons was higher than non-social icons, F(1,59) = 4.360, p = 0.041, η2 = 0.069. While the main effect of type of stimuli was non-significant, F(1, 59) = 0.107, p > 0.05, ηp2 = 0.002.
Direct comparisons between scores of attentional bias and 0 were performed in each group. In the PSMU/high FoMO group, scores of attentional bias was significantly lower than 0 only for social media icons, t(28) = -2.520, p = 0.018, Cohen's d = 0.47, but not for non-social icons, t(28) = 1.497, p > 0.05, Cohen's d = 0.28. In the PSMU/low FoMO group, scores of attentional bias was significantly higher than 0 only for social media icons, t(31) = 3.035, p = 0.005, Cohen's d = 0.54, but not for non-social icons, t(31) = 0.884, p > 0.05, Cohen's d = 0.16. These results indicated that the PSMU/high FoMO group had an attentional vigilance to social media icons, and the PSMU/low FoMO group had an attentional deceleration to social media icons.
Scores of attentional bias of first fixation duration
Repeated measures ANOVA for the scores of attentional bias of first fixation duration showed a main effect for type of stimuli, F(1, 59) = 6.023, p = 0.017, ηp2 = 0.093, indicating that the scores of attentional bias of first fixation duration were higher for non-social icons. While the main effect of group, F(1, 59) = 0.507, p > 0.05, ηp2 = 0.009, and the interaction between group and type of stimuli, F(1, 59) = 0.012, p > 0.05, ηp2 = 0.000, were both non-significant.
Scores of attentional bias of dwell time
Repeated measures ANOVA for the scores of attentional bias of dwell time showed an interaction effect between group and type of stimuli, F(1, 59) = 5.962, p = 0.018, ηp2 = 0.092. Simple effects analyses found higher scores of attentional bias towards social media icons in the PSMU/high FoMO group compared to the PSMU/low FoMO group, F(1,59) = 6.507, p = 0.013, η2 = 0.099. While the main effect of group, F(1, 59) = 3.380, p > 0.05, ηp2 = 0.054, and the main effect of type of stimuli, F(1, 59) = 0.037, p > 0.05, ηp2 = 0.001, were both non-significant.
Direct comparisons between scores of attentional bias and 50.0% were performed in each group. In the PSMU/high FoMO group, scores of attentional bias were significantly higher than 50.0% only for social media icons, t(28) = 2.074, p = 0.047, Cohen's d = 0.50, but not for non-social icons, t(28) = 0.865, p > 0.05, Cohen's d = 0.16. In the PSMU/low FoMO group, scores of attentional bias were not significantly higher than 50.0% for social media icons, t(31) = −1.239, p > 0.05, Cohen's d = 0.24, and non-social icons, t(31) = 0.496, p > 0.05, Cohen's d = 0.09. These results indicated that the PSMU/high FoMO group had an attentional maintenance to social media icons (Fig. 4).
Correlation Analyses. The correlational results indicated that the correlation trends between BSMAS scores, FoMO scores, and scores of attentional bias were different for different groups, supporting the notion that there might be different attentional bias patterns for different groups when faced with the same stimuli. Specifically, correlation analysis between BSMAS scores and FoMO scores showed a positively relationship in the PSMU/high FoMO group (r = 0.578, p = 0.001), but no significant relationship in PSMU/low FoMO group (r = 0.076, p = 0.679). Fisher's Z test found a significant difference between two correlation coefficients (p = 0.015). Correlation analysis between BSMAS scores and the scores of attentional bias of first fixation latency in social media icons showed a negative relationship in the PSMU/high FoMO group (r = −0.374, p = 0.045), but no significant relationship in PSMU/low FoMO group (r = −0.185, p = 0.310). Fisher's Z test found no significant difference between two correlation coefficients (p = 0.223). Correlation analysis between FoMO scores and the scores of attentional bias of first fixation latency in social media icons both showed a negative relationship in the PSMU/high FoMO group (r = −0.383, p = 0.040), and PSMU/low FoMO group (r = −0.385, p = 0.030). Fisher's Z test found no significant difference between two correlation coefficients (p = 0.496) (Fig. 5).
Discussion for Experiment 2
Experiment 2 found that the PSMU/high FoMO group showed attentional vigilance and maintenance toward social media icons, while the PSMU/low FoMO group had an attentional deceleration to social media icons, which supported hypothesis 2. This finding suggests that FoMO may have a kind of anxiety that impacts attentional bias patterns for PSMU (Przybylski et al., 2013); as prioritized attention was triggered when high FoMO faced social media icons to satisfy needs for accessing social information. Also, the correlational results indicated that correlation trends between BSMAS scores and scores of attentional bias were different between the high FoMO and low FoMO PSMU groups, supporting the idea that FoMO may be an important factor in influencing attentional bias patterns of individuals with PSMU.
General discussion
The present study investigated how FoMO moderated the attentional bias among PSMU by eye-movement experiments and focused on two issues: (a) whether individuals with PSMU have a specific attentional bias to social media icons differing from the control group; and (b) whether different FoMO levels in individuals with PSMU showed different attentional bias pattens toward social media icons. The results revealed that PSMU had an attentional maintenance to social media icons, and high FoMO PSMU individuals had attentional vigilance and maintenance toward social media icons. These results support previous research reporting similar results (Nikolaidou et al., 2019) and the I-PACE model (Brand et al., 2019).
Regarding the visual probe task combined with eye-movement, individuals with PSMU showed attentional maintenance toward social media icons, which is consistent with previous research (Nikolaidou et al., 2019). One potential reason for this finding could be that individuals with PSMU may have a lower cognitive flexibility compared with the control group (Jiang, Bai, et al., 2019), in that slower attention processing of PSMU when social media icons appeared. And, PSMU may also involve impaired attentional control (Brand et al., 2016, 2019; Wegmann, Müller, Turel, & Brand, 2020), manifested in attentional maintenance toward social media icons. Meanwhile, the correlational results indicated that correlation trends between BSMAS scores and scores of attentional bias were different between the PSMU and control groups, which may suggest that PSMU involves a specific attentional bias toward social media icons to a certain extent. However, Jiang et al. found that individuals with PSMU showed attentional vigilance and maintenance toward social text (Jiang, Bai, et al., 2019), and results of Experiment 1 indicated that PSMU only involved attentional maintenance toward social media icons. It suggests that type of stimuli may influence attentional bias among individuals with PSMU, which was similar to substance addiction (Wang et al., 2022), but further studies are necessary to support the similarity between PSMU and substance addiction in this part.
The results from Experiment 1 support the I-PACE model to a certain extent. Based on I-PACE, use of social media may meet social needs, then triggering an impulse to use social media under reinforcement and involving attentional bias towards the social stimuli (Brand et al., 2016, 2019). As the frequency of social media use increases, a cognitive bias like FoMO, may develop into stable PSMU under the compensation effect (a desire that the anxiety of missing out information can only alleviate through a consistent use of social media, Brand et al., 2016; Brand et al., 2019). Further, FoMO is an anxiety that others might be having rewarding experiences and want to stay connected continually with them (Przybyski et al., 2013). So, it may play a moderated role on the attentional bias among individuals with PSMU (Brand et al., 2016, 2019).
It is worth noting that the PSMU/high FoMO group showed an attentional bias in pattern of attentional vigilance and maintenance in social media icons, while the PSMU/low FoMO group showed an attentional deceleration to social media icons. It suggests that FoMO played a moderating role on the early components of attentional bias among those with PSMU, which was similar to anxiety (Zhao et al., 2023). Because of the anxiety of missing information about others, high FoMO users developed a seeking impulse to relevant stimuli (Robinson & Berridge, 2003) and showed attentional vigilance to the social media icons. Also, PSMU may produce a maintenance response to the social media icons due to the cognitive bias of social media use at a later stage because of the alleviation of negative emotions (Brand et al., 2019). Further, the correlational results indicated that correlation trends between BSMAS scores and scores of attentional bias were different between the PSMU/high FoMO and PSMU/low FoMO group, supporting the idea that high FoMO may lead to a more stable attentional bias toward social media icons among those with PSMU. Results of Experiment 2 indicated that FoMO may be an important factor in the development of PSMU (Fioravanti et al., 2021), which is supported by the I-PACE model (Brand et al., 2019).
The findings of this study should be considered in the context of its limitations. Firstly, the visual dot probe task has been a frequently used task to assess processes of attentional bias (Chapman, Devue, & Grimshaw, 2019). However, it has been associated with poor reliability and conflicting results recently (Thigpen, Gruss, Garcia, Herring, & Keil, 2018). Although this study assessed attentional bias among individuals with PSMU through a combined analysis of eye movement indicators and RT, future research can use the reverse correlation approach in combination with frequency-tagged steady-state visual potentials (SSVEPs) to improve the assessment confidence of the visual dot probe task. Secondly, different components of attentional bias can be computed through formulas (Kou et al., 2015; MacLeod & Mathews, 1988). However, the classical dot-probe task of lack of neutral-neutral pairs may decrease the possibility of finding attentional disengagement effects (Koster, Crombez, Verschuere, & Houwer, 2006). Future research can apply a revised dot-probe paradigm with additional neutral-neutral pairs to avoid this problem (Koster et al., 2006). Thirdly, criteria of PSMU included BSMAS item ratings of 3 or above on at least four of the six items (Andreassen et al., 2012), and also no structured clinical interview was conducted to determine whether participants met the inclusion and exclusion criteria in this study, which may not be rigorous enough. Social media is used in daily life widely, so PSMU may not be selected by an easy criterion, and future studies can use strict criteria to recruit participants with other BSMAS scale cutoff scores (Luo et al., 2021), further coupled with the structured clinical interview to recruit more qualified participants. Fourthly, though each pair of icons was created with color and complexity considered, no objective assessment was conducted. This may have a potential impact on attentional bias because of the different properties. Future studies should assess the complexity/salience of stimuli in an objective way to reduce the potential risk of influencing the results. Furthermore, there was a large discrepancy in the number of males and females. Although there were no gender differences in this study, future studies should consider the issue of gender balance.
In conclusion, the two experiments found that FoMO moderated the attentional bias among individuals PSMU, attentional maintenance toward social media icons was found among those with PSMU, and high FoMO individuals with PSMU had attentional vigilance and maintenance toward social media icons. Based on the I-PACE model, attentional bias toward social media icons among PSMU may be associated with reduced cognitive flexibility (Jiang, Bai, et al., 2019), which may result in more attentional resources allocated to social related stimuli (Jiang, Bai, et al., 2019). It has also been found that individuals with higher levels of FoMO suffered from impaired attentional control and may have difficulty in shifting attention away from socially relevant information (Yin et al., 2023), resulting in a more stable attentional bias towards social media icons among PSMU/high FoMO. Moreover, the gratification of PSMU/high FoMO of their sustained social needs through social media use may develop the levels of PSMU in turn (Brand et al., 2016, 2019), which means pay attention to the level of FoMO or attentional bias modification may increase the level of attention control and potentially reduce the development of PSMU in some degree. These results show that FoMO is an important factor on the development of PSMU and supports the views of I-PACE model. On the one hand, attentional bias toward social media icons can be sensitive markers of PSMU to characterize it as an addictive disorder. On the other hand, FoMO may be a cognitive mechanism on the development of PSMU, which suggests that it may be helpful to treat or intervene in PSMU through changing the irrational emotional cognition of FoMO and attentional bias training.
Funding sources
This study was supported by a grant from NSFC (grant number 32271140). We would also like to thank the participants who shared their experiences with us and participated in this study.
Authors' contribution
Each author's contribution to the manuscript like this: YW – investigation, analysis and interpretation of data, statistical analysis; JD. E − study supervision; CM – study supervision; LZ – investigation, statistical analysis; HB. Y– study concept and design, study supervision. All authors had full access to all data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Conflict of interest
The authors declare no conflict of interest.
Supplementary materials
Supplementary data to this article can be found online at https://doi.org/10.1556/2006.2024.00039.
References
Altuwairiqi, M., Jiang, N., & Ali, R. (2019). Problematic attachment to social media: Five behavioural archetypes. International Journal of Environmental Research and Public Health, 16(12), 2136. https://doi.org/10.3390/ijerph16122136.
American Psychiatric Association (APA) (2013). Diagnostic and statistical manual of mental disorders DSM-5 (5th ed.). American Psychiatric Publishing.
Anderson, B. A. (2016). What is abnormal about addiction-related attentional biases? Drug and Alcohol Dependence, 167, 8–14. https://doi.org/10.1016/j.drugalcdep.2016.08.002.
Andreassen, C. S., Torsheim, T., Brunborg, G. S., & Pallesen, S. (2012). Development of a Facebook addiction scale. Psychological Reports, 110(2), 501–517. https://doi.org/10.2466/02.09.18.PR0.110.2.501-517.
Bandura, A. (2001). Social cognitive theory of mass communication. Media Psychology, 17, 265–299. https://doi.org/10.1207/S1532785XMEP0303_03.
Bielefeld, M., Drews, M., Putzig, I., Bottel, L., Steinbüchel, T., Dieris-Hirche, J., … Theodor Te Wildt, B. (2017). Comorbidity of Internet use disorder and attention deficit hyperactivity disorder: Two adult case-control studies. Journal of Behavioral Addictions, 6(4), 490–504. https://doi.org/10.1556/2006.6.2017.073.
Brand, M., Rumpf, H., Demetrovics, Z., Müller, A., Stark, R., King, D. L., … Potenza, M. N. (2022). Which conditions should be considered as disorders in the International Classification of Diseases (ICD-11) designation of “other specified disorders due to addictive behaviors”. Journal of Behavioral Addictions, 11(2), 150–159. https://doi.org/10.1556/2006.2020.00035.
Brand, M., Wegmann, E., Stark, R., Müller, A., Wölfling, K., Robbins, T. W., & Potenza, M. N. (2019). The Interaction of Person-Affect-Cognition-Execution (I-PACE) model for addictive behaviors: Update, generalization to addictive behaviors beyond internet-use disorders, and specification of the process character of addictive behaviors. Neuroscience and Biobehavioral Reviews, 104, 1–10. https://doi.org/10.1016/j.neubiorev.2019.06.032.
Brand, M., Young, K. S., Laier, C., Wölfling, K., & Potenza, M. N. (2016). Integrating psychological and neurobiological considerations regarding the development and maintenance of specific Internet-use disorders: An Interaction of Person-Affect-Cognition-Execution (I-PACE) model. Neuroscience & Biobehavioral Reviews, 71, 252–266. https://doi.org/10.1016/j.neubiorev.2016.08.033.
Chapman, A., Devue, C., & Grimshaw, G. M. (2019). Fleeting reliability in the dot-probe task. Psychological Research, 83, 308–320. https://doi.org/10.1007/s00426-017-0947-6.
Davis, R. A. (2001). A cognitive–behavioral model of pathological Internet use. Computers in Human Behavior, 17(2), 187–195. https://doi.org/10.1016/S0747-5632(00)00041-8.
Elhai, J. D., Yang, H., & Montag, C. (2021a). Fear of missing out (FOMO): Overview, theoretical underpinnings, and literature review on relations with severity of negative affectivity and problematic technology use. Revista brasileira de psiquiatria (Sao Paulo, Brazil: 1999), 43(2), 203–209. https://doi.org/10.1590/1516-4446-2020-0870.
Eysenck, M. W., Derakshan, N., Santos, R., & Calvo, M. G. (2007). Anxiety and cognitive performance: Attentional control theory. Emotion (Washington, D.C.), 7(2), 336–353. https://doi.org/10.1037/1528-3542.7.2.336.
Field, M., & Cox, W. M. (2008). Attentional bias in addictive behaviors: A review of its development, causes, and consequences. Drug and Alcohol Dependence, 97(1–2), 1–20. https://doi.org/10.1016/j.drugalcdep.2008.03.030.
Field, M., Werthmann, J., Franken, I., Hofmann, W., Hogarth, L., & Roefs, A. (2016). The role of attentional bias in obesity and addiction. Health Psychology: Official Journal of the Division of Health Psychology, American Psychological Association, 35(8), 767–780. https://doi.org/10.1037/hea0000405.
Fioravanti, G., Casale, S., Benucci, S. B., Prostamo, A., Falone, A., Ricca, V., & Rotella, F. (2021). Fear of missing out and social networking sites use and abuse: A meta-analysis. Computers in Human Behavior, 122, 106839. https://doi.org/10.1016/J.CHB.2021.106839.
Fumagalli, E., Dolmatzian, M. B., & Shrum, L. J. (2021). Centennials, FOMO, and loneliness: An investigation of the impact of social networking and messaging/VoIP apps usage during the initial stage of the coronavirus pandemic. Frontiers in Psychology, 12, 620739. https://doi.org/10.3389/fpsyg.2021.620739.
Gao, X. L., Li, T. Z., Bai, X. J., & Gao, L. (2023). Effect of self-control trait on attentional bias of food cues and its intervention: Evidence from eye movements. Journal of Psychological Science, (01), 25–33. https://doi.org/10.16719/j.cnki.1671-6981.20230104.
Hou, J., Zhu, Y. G., & Fang, X. Y. (2021). Mobile phone addiction and depression: Multiple mediating effects of social anxiety and attentional bias to negative emotional information. Acta Psychologica Sinica, (04), 362–373. https://doi.org/10.3724/SP.J.1041.2021.00362.
Jiang, Y. Z., Bai, H. M., Bai, X. L., Qi, S. S., & Wu, Y. T. N. (2019). Research on characteristics of attentional bias in the problematic mobile social network use behavior. Chinese Journal of Drug Dependence, (05), 353–358. https://doi.org/10.13936/j.cnki.cjdd1992.2019.05.006.
Jiang, J. Q., Wang, H. U., & Qiang, M. Y. (2019). Dynamic attention bias in social anxiety. Advances in Psychological Science, 27(11), 1887–1895. https://doi.org/10.3724/SP.J.1042.2019.01887.
Jones, A., Worrall, S., Rudin, L., Duckworth, J. J., & Christiansen, P. (2021). May I have your attention, please? Methodological and analytical flexibility in the addiction stroop. Addiction Research & Theory, 29(5), 413–426. https://doi.org/10.1080/16066359.2021.1876847.
Koster, E. H. W., Crombez, G., Verschuere, B., & Houwer, J. D. (2006). Attention to threat in anxiety-prone individuals: Mechanisms underlying attentional bias. Cognitive Therapy and Research, 30(5), 635–643. https://doi.org/10.1007/s10608-006-9042-9.
Kou, H., Su, Y. H., Luo, X. C., & Chen, H. (2015). Attentional bias toward face-related words among females with facial negative physical self: Evidence from an eye-movement study. Acta Psychologica Sinica, 47(10), 1213–1222. https://doi.org/10.3724/SP.J.1041.2015.01213.
Kuss, D. J., & Griffiths, M. D. (2017). Social networking sites and addiction: Ten lessons learned. International Journal of Environmental Research and Public Health, 14(3), 311. https://doi.org/10.3390/ijerph14030311.
Luo, T., Qin, L., Cheng, L., Wang, S., Zhu, Z., Xu, J., … Liao, Y. (2021). Determination the cut-off point for the Bergen social media addiction (BSMAS): Diagnostic contribution of the six criteria of the components model of addiction for social media disorder. Journal of Behavioral Addictions, 10(2), 281–290. https://doi.org/10.1556/2006.2021.00025.
MacLeod, C., & Mathews, A. (1988). Anxiety and the allocation of attention to threat. The Quarterly Journal of Experimental Psychology. A, Human Experimental Psychology, 40(4), 653–670. https://doi.org/10.1080/14640748808402292.
MacLeod, C., Mathews, A., & Tata, P. (1986). Attentional bias in emotional disorders. Journal of Abnormal Psychology, 95(1), 15–20. https://doi.org/10.1037//0021-843x.95.1.15.
Matias, J., Quinton, J. C., Colomb, M., Normand, A., Izaute, M., & Silvert, L. (2021). Fear of missing out predicts distraction by social reward signals displayed on a smartphone in difficult driving situations. Frontiers in Psychology, 12, 688157. https://doi.org/10.3389/fpsyg.2021.688157.
Miranda, S., Trigo, I., Rodrigues, R., & Duarte, M. (2023). Addiction to social networking sites: Motivations, flow, and sense of belonging at the root of addiction. Technological Forecasting and Social Change, 188. https://doi.org/10.1016/j.techfore.2022.122280.
Montag, C., Wegmann, E., Sariyska, R., Demetrovics, Z., & Brand, M. (2021). How to overcome taxonomical problems in the study of Internet use disorders and what to do with “smartphone addiction”. Journal of Behavioral Addictions, 9(4), 908–914. https://doi.org/10.1556/2006.8.2019.59.
Nasser, N. S., Sharifat, H., Rashid, A. A., Hamid, S. A., Rahim, E. A., Loh, J. L., … Suppiah, S. (2020). Cue-reactivity among young adults with problematic Instagram use in response to Instagram-themed risky behavior cues: A pilot fMRI study. Frontiers in Psychology, 11, 556060. https://doi.org/10.3389/fpsyg.2020.556060.
Nikolaidou, M., Fraser, D. S., & Hinvest, N. (2019). Attentional bias in Internet users with problematic use of social networking sites. Journal of Behavioral Addictions, 8(4), 733–742. https://doi.org/10.1556/2006.8.2019.60.
Opsenica Kostić, J., Pedović, I., & Stošić, M. (2022). Predicting social media use intensity in late adolescence: The role of attachment to friends and fear of missing out. Acta Psychologica, 229, 103667. https://doi.org/10.1016/j.actpsy.2022.103667.
Pabst, A., Bollen, Z., Masson, N., Billaux, P., de Timary, P., & Maurage, P. (2023). An eye-tracking study of biased attentional processing of emotional faces in severe alcohol use disorder. Journal of Affective Disorders, 323, 778–787. https://doi.org/10.1016/j.jad.2022.12.027.
Przybylski, A. K., Murayama, K., DeHaan, C. R., & Gladwell, V. (2013). Motivational, emotional, and behavioral correlates of fear of missing out. Computers in Human Behavior, 29(4), 1841–1848. https://doi.org/10.1016/j.chb.2013.02.014.
Robinson, T. E., & Berridge, K. C. (2003). Addiction. Annual Review of Psychology, 54, 25–53. https://doi.org/10.1146/annurev.psych.54.101601.145237.
Sha, P., Sariyska, R., Riedl, R., Lachmann, B., & Montag, C. (2019). Linking internet communication and smartphone use disorder by taking a closer look at the Facebook and WhatsApp applications. Addictive behaviors reports, 9, 100148. https://doi.org/10.1016/j.abrep.2018.100148.
Sheppes, G., Luria, R., Fukuda, K., & Gross, J. J. (2013). There’s more to anxiety than meets the eye: Isolating threat-related attentional engagement and disengagement biases. Emotion, 13(3), 520–528. https://doi.org/10.1037/a0031236.
Soh, P. C., Charlton, J. P., & Chew, K. (2014). The influence of parental and peer attachment on Internet usage motives and addiction. First Monday, 19. http://www.ojphi.org/ojs/index.php/fm/article/view/5099.
Statista (2024). Social media usage worldwide. https://www.statista.com/statistics/278414/number-of-worldwide-social-network-users [accessed 16 Feb 2024].
Su, C., Zhou, H., Gong, L., Teng, B., Geng, F., & Hu, Y. (2021). Viewing personalized video clips recommended by TikTok activates default mode network and ventral tegmental area. NeuroImage, 237, 118136. https://doi.org/10.1016/j.neuroimage.2021.118136.
Thigpen, N. N., Gruss, L. F., Garcia, S., Herring, D. R., & Keil, A. (2018). What does the dot-probe task measure? A reverse correlation analysis of electrocortical activity. Psychophysiology, 55(6), e13058. https://doi.org/10.1111/psyp.13058.
Thomson, K., Hunter, S. C., Butler, S. H., & Robertson, D. J. (2021). Social media ‘addiction’: The absence of an attentional bias to social media stimuli. Journal of Behavioral Addictions, 10(2), 302–313. https://doi.org/10.1556/2006.2021.00011.
Wang, Y., Liu, B., Zhang, L., & Zhang, P. (2022). Anxiety, depression, and stress are associated with internet gaming disorder during COVID-19: Fear of missing out as a mediator. Frontiers in Psychiatry, 13, 827519. https://doi.org/10.3389/fpsyt.2022.827519.
Wang, F. X., Wang, Y. Q., & Hao, Y. B. (2022). Attention bias to drug related cues of female heroin abstainer: Evidence from eye movements. Journal of Psychological Science(02), 481–490. https://doi.org/10.16719/j.cnki.1671-6981.20220228.
Wang, P., Xie, X., Wang, X., Wang, X., Zhao, F., Chu, X., … Lei, L. (2018). The need to belong and adolescent authentic self-presentation on SNSs: A moderated mediation model involving FoMO and perceived social support. Personality and Individual Differences, 128, 133–138. https://doi.org/10.1016/j.paid.2018.02.035.
Wegmann, E., Müller, S. M., Turel, O., & Brand, M. (2020). Interactions of impulsivity, general executive functions, and specific inhibitory control explain symptoms of social-networks-use disorder: An experimental study. Scientific Reports, 10(1), 3866. https://doi.org/10.1038/s41598-020-60819-4.
WHO (2023). International Classification of Diseases (11th ed., [ICD-11]). https://www.who.int/lassifications/icd/en [accessed 23 Jan 2023].
Xu, Y., Chen, Q., & Tian, Y. (2024). The impact of problematic social media use on inhibitory control and the role of fear of missing out: Evidence from event-related potentials. Psychology Research and Behavior Management, 17, 117–128. https://doi.org/10.2147/PRBM.S441858.
Yin, Y., Cai, X., Ouyang, M., Li, S., Li, X., & Wang, P. (2023). FoMO and the brain: Loneliness and problematic social networking site use mediate the association between the topology of the resting-state EEG brain network and fear of missing out. Computers in Human Behavior, 141, 107624. https://doi.org/10.1016/j.chb.2022.107624.
Zhao, R., Ma, W., Li, C., Yang, M., He, S., Mao, N., … Cui, L. (2023). Trait anxiety is related to an impaired attention model for controllable threat cues: Evidence from ERPs. Biological Psychology, 177, 108508. https://doi.org/10.1016/j.biopsycho.2023.108508.
Zhao, J., Zhou, Z., Sun, B., Zhang, X., Zhang, L., & Fu, S. (2022). Attentional bias is associated with negative emotions in problematic users of social media as measured by a dot-probe task. International Journal of Environmental Research and Public Health, 19(24), 16938. https://doi.org/10.3390/ijerph192416938.