The mediating role of phubbing in the relationship between self-efficacy and problematic Instagram use

Abstract

This study explores the mediating role of phubbing behaviour in the relationship between university students’ self-efficacy and problematic Instagram use. Utilizing a quantitative relational design, Structural Equation Modeling (SEM) was employed to analyse the relationships among the variables. The findings reveal an association where higher self-efficacy is related to lower problematic Instagram use. This relationship appears to be mediated by phubbing behaviours, with higher self-efficacy also being associated with lower phubbing. Higher self-efficacy was associated with lower phubbing tendencies and lower reported problematic Instagram use. The study underscores the importance of addressing both psychological traits, such as self-efficacy, and behavioural factors, like phubbing, in developing interventions to address problematic social media use. Recommendations include implementing self-efficacy enhancement programs and raising awareness of the detrimental associations of phubbing on university campuses. These results contribute to a deeper understanding of the interplay between psychological and behavioural factors in problematic social media use and offer potential considerations for prevention efforts.

Introduction

The widespread use of smartphones and social media platforms has significantly altered individuals’ interactions with one another and their self-perceptions. This transformation has given rise to behaviours such as “phubbing” and social media addiction, which have become increasingly prevalent among young adults. Phubbing, defined as the act of ignoring a conversation partner in a social setting by focusing on one’s phone instead [1, 2], has been associated with interpersonal conflicts, weakened emotional bonds, and general dissatisfaction within relationships [2]. Similarly, PIU often involves excessive use of the platform and has been linked to adverse impacts on daily life [3, 4]. Research indicates that PIU is particularly prevalent among young adults and has been linked to negative consequences for psychological well-being, social relationships, and academic performance [5, 6]. This issue is especially pertinent in contexts like Türkiye, which reports one of the highest rates of social media usage globally, with a significant portion of its young population actively engaged on platforms like Instagram [7]. The unique cultural and social dynamics in Türkiye may also influence how these digital behaviours manifest and interact with psychological traits, making it a critical region for such research.

Psychological characteristics are also considered important in relation to these behaviours. Self-efficacy, a fundamental psychological construct, refers to individuals’ belief in their ability to cope with challenging situations and achieve their goals [8]. Individuals reporting high self-efficacy also tend to report being better equipped to manage stress and often display a more positive attitude towards difficulties encountered in life [9, 10]. However, the relationship between self-efficacy and digital behaviours remains insufficiently explored.

While the negative association between self-efficacy and various addictive behaviours is documented [5, 11], and the link between phubbing and negative social outcomes is established [1, 2, 12], the interplay between these three specific constructs—self-efficacy, phubbing, and PIU—remains underexplored. Previous research has often examined these variables in separate pairs: for instance, the link between PIU and phubbing [13, 14], or between self-efficacy and general internet addiction [5]. However, the literature lacks a specific examination of how self-efficacy relates to PIU through the mechanism of phubbing.

This study, therefore, aims to fill this critical gap by proposing and testing a mediational model where phubbing acts as the intermediary mechanism linking self-efficacy to PIU. By investigating this pathway, we seek to answer how a key psychological resource (self-efficacy) might be associated with a specific problematic digital behaviour (PIU), with phubbing serving as the explanatory behavioural link. Furthermore, grounding this investigation in a sample of university students in Türkiye addresses a call for more research in non-Western contexts, where cultural norms surrounding technology use and social interaction may uniquely shape these relationships. Understanding this dynamic is particularly pertinent for university students, a demographic highly engaged with social media and susceptible to its problematic aspects. This research, therefore, seeks to fill this gap by providing empirical evidence on how a prevalent digital behaviour (phubbing) may explain the link between a key psychological resource (self-efficacy) and a contemporary mental health concern PIU in this specific population. Furthermore, by focusing on university students in Türkiye, this research sheds light on the digital behaviours and psychological characteristics of this group, offering insights that could inform culturally relevant prevention and intervention strategies. The use of structural equation modelling provides a robust methodological approach to uncovering the relationships between these variables. This study’s correlational framework also ensures a realistic interpretation of the findings, avoiding unwarranted causal claims. The conceptual framework of this research posits that self-efficacy is associated with PIU both directly and indirectly through phubbing behaviours. These findings could enhance our understanding of the psychological underpinnings of digital behaviours and contribute to the development of interventions aimed at addressing adverse associations of excessive social media use.

Present study and hypotheses

The conceptual framework of this research posits that self-efficacy is associated with PIU both directly and indirectly through phubbing behaviours. Based on the existing literature and our theoretical framework, we formulated the following hypotheses:

• H1: Self-efficacy will be negatively associated with problematic Instagram use (PIU).

• H2: Self-efficacy will be negatively associated with phubbing.

• H3: Phubbing will be positively associated with problematic Instagram use (PIU).

• H4: Phubbing will mediate the relationship between self-efficacy and PIU, such that there will be a significant indirect association from self-efficacy to PIU through phubbing.

Phubbing

In recent years, with the widespread adoption of smartphones, a behaviour known as “phubbing” has emerged as a phenomenon significantly affecting social interactions between individuals. Phubbing is defined as the act of ignoring a person in a social setting by focusing on one’s phone instead [1, 2]. This behaviour has been associated with negative outcomes, particularly in familial relationships, romantic partnerships, and professional settings. For instance, parental phone use during interactions with children has been linked to children’s social withdrawal behaviours and an increase in parent-child conflict [15].

Phubbing was first introduced in academic literature in 2012 and has since become a widely researched topic, especially in the fields of psychology, communication studies, and education [1]. This behaviour has been reported to have significant implications for individuals’ emotional well-being, being associated with psychological issues such as depression, anxiety, and loneliness [2]. Furthermore, studies have demonstrated associations where phubbing is linked to negative impacts on romantic relationships and is often observed alongside trust issues and relationship dissatisfaction [2].

Research on the causes and consequences of phubbing indicates that this behaviour can be viewed both as a form of technological addiction and a social phenomenon [2]. The constant accessibility of smartphones and individuals’ growing dependence on these devices play a critical role in the prevalence of phubbing. Additionally, how phubbing is perceived within a social context and its impact on those affected have also garnered significant attention from researchers [16].

Self-efficacy

General self-efficacy refers to individuals’ beliefs in their ability to cope with various challenges and achieve their goals, making it a crucial psychological construct. According to Bandura’s social cognitive theory, self-efficacy is a fundamental factor that shapes individuals’ behaviours, motivations, and emotional responses. It encompasses beliefs about overcoming difficulties and achieving objectives, which are thought to be significantly related to performance, perseverance, and stress management abilities [9, 17].

General self-efficacy is particularly significant in the context of education and academic achievement. Research indicates that students with higher self-efficacy beliefs often report better academic performance, higher motivation during the learning process, and demonstrate greater resilience when facing challenges [9, 18]. Self-efficacy is also associated with students’ ability to select learning strategies, set goals, and pursue these goals effectively. For instance, students with higher self-efficacy are more likely to set ambitious goals and exert greater effort to achieve them [18].

Beyond academic success, general self-efficacy is closely linked to individuals’ psychological well-being and overall life satisfaction. Individuals with strong self-efficacy beliefs are better equipped to manage stress and maintain a positive attitude toward life’s challenges [10]. Higher self-efficacy, associated with enhanced confidence in abilities, is often linked to addressing difficulties more effectively. This is particularly important for individuals in transitional stages, such as university students [9, 10].

General self-efficacy is also associated with other psychological constructs, such as character strengths and hope. For example, character strengths like leadership, hope, and perseverance may be related to higher self-efficacy, which, in turn, shows a positive association with academic self-efficacy [9]. These findings underscore the critical role of self-efficacy in shaping individuals’ personal and academic lives.

Problematic Instagram use (PIU)

PIU has emerged as a growing issue, particularly among young adults and adolescents in recent years. This addiction is characterised by the excessive use of Instagram, which is often linked to negative impacts on individuals’ daily lives. Instagram’s reward mechanisms, such as constant notifications, likes, and comments, are designed to effectively capture users’ attention and may be associated with addictive-like behaviours [3, 4]. PIU is significantly associated with psychological well-being, social relationships, and academic performance [5, 6].

The design and algorithms of the platform are considered to play a crucial role in relation to PIU. Specifically, the personalised content tailored to users’ interests and the continuously updated feed encourage prolonged usage [19, 20]. Additionally, according to social comparison theory, Instagram users may negatively evaluate their own lives when observing the “perfect” lifestyles shared by others; this pattern has been associated with addictive-like behaviour [4, 21].

PIU has been linked to profound psychological and social consequences. Psychological effects include depression, anxiety, and low self-esteem [3, 4]. Extended periods spent on Instagram have been associated with reduced real-life social interactions and a weakening of societal bonds [20]. Furthermore, PIU has been associated with declining academic performance and reduced productivity, particularly among younger individuals [3, 22].

A negative relationship exists between PIU and life satisfaction. Studies indicate an association where higher PIU is linked with heightened feelings of loneliness, which in turn are related to diminished life satisfaction [21]. During the COVID-19 pandemic, increased Instagram usage was associated with intensified feelings of loneliness and was adversely related to users’ life satisfaction [21]. The need for continuous validation through likes and comments on Instagram may be associated with individuals’ self-esteem becoming more dependent on external factors, a pattern often observed alongside addictive-like behaviours [20, 21].

Various strategies have been proposed to mitigate PIU. These include limiting the time spent on Instagram, enhancing real-life social interactions, and fostering self-esteem [19, 20]. Additionally, modifications to Instagram’s design and raising awareness about the platform’s negative impacts may be associated with a reduction in addictive-like behaviours [19, 23].

PIU is associated with concepts such as self-efficacy and phubbing. General self-efficacy refers to individuals’ belief in their ability to cope with challenges and achieve goals. Those reporting high self-efficacy also tend to report being more adept at managing stress and often exhibit a more positive outlook on life’s difficulties [5, 24]. Conversely, phubbing, defined as ignoring others in social settings in favour of one’s smartphone, has been linked to detrimental consequences for family, romantic, and workplace relationships [25, 13, 14].

Research demonstrates an association where higher PIU is related to intensified feelings of loneliness and, subsequently, lower life satisfaction [5]. Moreover, PIU has been negatively associated with self-esteem, particularly when individuals perceive their own lives as inadequate in comparison to the “ideal” lives portrayed on the platform [26]. Phubbing behaviour, which may be associated with excessive use of platforms like Instagram, is linked with reduced real-life social interactions and a weakening of communal ties [13, 14]. Additionally, phubbing has been associated with adverse outcomes in romantic relationships, such as an erosion of trust and lower relationship satisfaction [24, 14].

This study aims to evaluate the mediating role of phubbing behaviours in the relationship between university students’ perceptions of self-efficacy and their PIU. The conceptual framework representing this study can be found in Fig. 1.

Fig. 1.

The conceptual framework

Method

Research design

In this study, a relational design, categorized under quantitative research methods, was employed to explore the potential direct or indirect associations among the variables under investigation. Quantitative data provide a broad framework for addressing the research problem [27]. To evaluate the proposed model, Structural Equation Modeling (SEM) was utilized. SEM refers to a statistical technique widely applied across various scientific disciplines [28].

This study, designed based on the relational model of quantitative research approaches, aims to evaluate the potential mediating role of phubbing in the relationship between self-efficacy and PIU among university students. Consistent with the nature of relational research, the study seeks to identify the role of the mediator variable (general phubbing) in understanding the association between the independent variable (self-efficacy) and the dependent variable (PIU).

Sample and data collection

The population for this study consisted of students currently studying in the Faculty of Education in Türkiye at the time of data collection. The non-probability sampling techniques of convenience and chain referral were used to recruit participants. The data of the study were collected in a digital environment. For this purpose, the measurement tools used in the research were recreated on Google Forms and sent to the study group. The sample included 310 female (75.6%) and 100 male (24.4%) participants, totalling 410 university students. The participants’ ages ranged from 18 to 55, with an average age of 23.30 (SD = 4.53).

Data collection tools

The data collection tools used in this study include the General Self-Efficacy Scale, the General Phubbing Scale, and the Instagram Addiction Scale (for PIU), as well as a personal information form to collect demographic data.

General Self-Efficacy scale

The measurement tool developed by Sherer et al. [29] to assess individuals’ general self-efficacy was adapted into Turkish by Yıldırım and İlhan [30]. The tool consists of 17 items and 3 subscales (Initiation, Persistence, and Effort to Maintain). High scores on this scale indicate a high level of general self-efficacy. The scale uses a 5-point Likert-type evaluation, with reverse-coded items (2, 4, 5, 6, 7, 10, 11, 12, 14, 16, and 17). The internal consistency coefficient for the entire scale was found to be 0.80. In this study, the internal consistency coefficient of the scale was found to be 0.87.

General phubbing scale

The measurement tool used to determine individuals’ phubbing behaviours was developed by Chotpitayasunondh and Douglas [31]. The Turkish adaptation of the tool was carried out by Ergün, Göksu, and Sakız [32]. The tool consists of 15 items and 4 subscales (Nomophobia, Interpersonal Conflict, Self-Isolation, and Problem Acknowledgment). The scale does not contain any reverse-scored questions. The 7-point Likert-type scale shows that low scores indicate low levels of phubbing, while high scores indicate high levels of phubbing. The internal consistency coefficient for the entire scale was found to be 0.91. In this study, the internal consistency coefficient of the scale was also found to be 0.91.

Instagram addiction scale

The measurement tool developed by D’Souza et al. [33] to assess Instagram addiction levels was adapted into Turkish by Kavaklı and İnan [34]. The scale consists of 21 items and 5 subscales (Lack of Control, Escapism, Disengagement, Obsession, Health, and Interpersonal Trouble). High scores on the scale indicate a high level of Instagram addiction. The internal consistency coefficient for the scale was found to be 0.92. In this study, the internal consistency coefficient of the scale was found to be 0.94.

Data analysis

In this study, which examines the mediating role of phubbing in the relationship between self-efficacy and PIU, preliminary analyses including missing data analysis, outlier analysis, normality distribution, and linearity analysis were conducted before proceeding with the statistical analyses (Table 1).

Table 1.

Kurtosis and skewness coefficients of measurement tools

Variables	Mean	Sd	Skewness	Kurtosis
Self-Efficacy	62,2481	10,59164	-,168	-,610
Initiative	32,9073	7,04643	-,514	,035
Effort	18,5387	3,46661	-,174	-,554
Persistence	10,8021	2,02911	-,038	-,231
PIU	41,8208	15,84383	,624	-,616
Lack of Control	6,9738	3,31920	1,122	,657
Escapism	9,6022	4,61237	,487	-,839
Disengagement	12,4321	5,66105	,684	-,467
Obsession	7,9981	3,20310	,630	-,319
Health and Interpersonal Trouble	4,8147	2,43845	1,330	,862
General Phubbing	40,5506	15,04319	,619	-,130
Nomophobia	15,7933	6,07105	,077	-,778
Interpersonal Conflict	8,1592	4,74597	1,334	1,399
Self-Isolation	8,0763	4,62280	1,248	,959
Problem Acknowledgment	8,5217	4,10804	,577	-,202

In the second phase, the relationships between the variables to be included in the structural equation modeling were calculated using the Pearson Product-Moment Correlation Coefficient. In the first stage of the third phase, prior to performing the structural equation analysis, the measurement model was first tested following the recommendations of Anderson and Gerbing [35] and Kline [28]. After validating the measurement model, the structural model was developed. In the second stage of the third phase, the four conditions for analysis according to the Baron and Kenny [36] method were considered, adapted for an associational interpretation. The first condition was to examine whether the independent variable (self-efficacy) is significantly associated with the dependent variable PIU. The second condition was to investigate whether the independent variable (self-efficacy) is significantly associated with the mediator variable (phubbing). The third condition was to examine whether the mediator variable (phubbing) is significantly associated with the dependent variable PIU. Finally, when the mediator variable (phubbing) is analysed along with the independent variable (self-efficacy), the association between the independent variable and the dependent variable should show a reduction in magnitude or become non-significant, while the mediator variable should still show a significant association with the dependent variable [36]. After verifying the stated conditions and testing the models, the bootstrapping method was used to determine whether the indirect effects were significant. The bootstrapping method is a mini trial of the population [37]. The bootstrapping method was conducted with 5000 resamples at a 95% confidence interval. For the indirect effect to be considered significant, the confidence intervals should not include zero [38].

Findings

Descriptive and correlational statistics of variables in the structural equation model

In the study, the relationships between the variables in the structural equation model were analysed using Pearson Product-Moment Correlation. The results of the Pearson Product-Moment Correlation analysis are presented in Table 2.

Table 2.

Correlations between variables

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
1	1
2	,930**	1
3	,846**	,639**	1
4	,545**	,292**	,487**	1
5	-,355**	-,324**	-,295**	-,223**	1
6	-,353**	-,313**	-,308**	-,231**	,825**	1
7	-,279**	-,261**	-,203**	-,206**	,825**	,570**	1
8	-,327**	-,286**	-,285**	-,224**	,937**	,746**	,714**	1
9	-,163**	-,148**	-,156**	-,070	,672**	,392**	,426**	,554**	1
10	-,323**	-,326**	-,250**	-,130**	,758**	,675**	,473**	,670**	,424**	1
11	-,307**	-,300**	-,233**	-,165**	,608**	,499**	,510**	,587**	,351**	,485**	1
12	-,157**	-,117*	-,168**	-,126*	,457**	,298**	,445**	,463**	,286**	,271**	,756**	1
13	-,324**	-,342**	-,208**	-,150**	,489**	,431**	,365**	,442**	,327**	,444**	,804**	,429**	1
14	-,258**	-,272**	-,169**	-,114*	,387**	,367**	,317**	,366**	,178**	,335**	,746**	,295**	,569**	1
15	-,228**	-,224**	-,173**	-,116*	,551**	,476**	,431**	,542**	,286**	,484**	,778**	,461**	,513**	,511**	1

Note: Pearson’s r correlation coefficients are presented. (1) Self-Efficacy; (2) Initiative; (3) Effort; (4) Persistence; (5) PIU; (6) Lack of Control; (7) Escapism; (8) Disengagement; (9) Obsession; (10) Health and Interpersonal Trouble; 11. General Phubbing; 12. Nomophobia; 13. Interpersonal Conflict; 14. Self-Isolation; 15. Problem Acknowledgment

*p < .01; p < .05

There were very strong positive and significant relationships between Self-Efficacy (1) and Initiative (2) (r = .930, p < .01) and Effort (3) (r = .846, p < .01). This situation shows that individuals’ self-confidence is closely related to their tendencies to take initiative and exert effort. In addition, there was a moderate-level positive and significant relationship between Persistence (4) and Self-Efficacy (1) (r = .545, p < .01).

There were negative significant relationships between PIU (5) and Self-Efficacy (1) (r = − .355, p < .01), Initiative (2) (r = − .324, p < .01), Effort (3) (r = − .295, p < .01) and Persistence (4) (r = − .223, p < .01). These results suggest a negative association between PIU and individuals’ self-reported self-confidence and effort-exerting characteristics. Strong positive relationships were observed between PIU sub-dimensions of Lack of Control (6), Escapism (7), and Disengagement (8) (Lack of Control and Escapism, r = .825, p < .01; Lack of Control and Disengagement, r = .746, p < .01). This situation shows that the addiction sub-dimensions are highly related to each other. Obsession (9) has positive significant relationships with the addiction sub-dimensions of Lack of Control (6) (r = .392, p < .01), Escapism (7) (r = .426, p < .01), and Disengagement (8) (r = .554, p < .01). Health and Interpersonal Trouble (10) has moderate-level positive significant relationships with addiction sub-dimensions (Lack of Control, r = .675, p < .01; Disengagement, r = .670, p < .01) and other addiction indicators such as Obsession (9) (r = .424, p < .01).

General Phubbing (11) has strong positive significant relationships with Nomophobia (12) (r = .756, p < .01), Interpersonal Conflict (13) (r = .804, p < .01) and Self-Isolation (14) (r = .746, p < .01). These findings highlight the relationship between constant phone use and interpersonal conflict and loneliness. Nomophobia (12) is related to addiction symptoms and dimensions that harm social relationships (Obsession, r = .286, p < .01; Interpersonal Conflict, r = .429, p < .01). There was a significant positive relationship between Interpersonal Conflict (13) and Problem Acknowledgment (15) (r = .513, p < .01). This finding shows that individuals’ awareness of their problems may increase their perception of interpersonal conflict. Problem Acknowledgment (15) also has positive significant relationships with Health and Interpersonal Trouble (10) (r = .484, p < .01) and Self-Isolation (14) (r = .511, p < .01).

Findings on the test of the measurement model

Before examining the structural equation model that investigating the mediating role of phubbing in the relationship between self-efficacy and PIU, as suggested by Anderson and Gerbing [35] and Kline [28], the measurement model was first tested. Once the measurement model was confirmed, the structural model was tested.

The measurement model in this study includes three latent variables: self-efficacy, phubbing, and PIU. A total of 12 observed variables correspond to these latent variables. The self-efficacy scale has three subdimensions, with each subdimension forming a parcel for the self-efficacy variable. The phubbing scale consists of four subdimensions, and each dimension represents a parcel for the phubbing variable. Similarly, the Instagram addiction scale comprises five subdimensions, with each dimension forming a parcel for the PIU variable.

The goodness-of-fit indices for the confirmatory factor analysis applied to the measurement model are presented in Table 3.

Table 3.

Goodness-of-fit indices for the measurement model

Goodness-of-fit measures	Good fit	Acceptable fit	Model Value	Comment
χ²			174.49
Sd			49
χ²/ df	0 ≤ χ²/df ≤ 2	2 < χ²/df ≤ 0.5	3.56	Acceptable fit
RMSEA	0.00 ≤ RMSEA < 0.05	0.05 ≤ RMSEA ≤ 0.08	0.079	Acceptable fit
SRMR	0.00 ≤ SRMR < 05	0.05 ≤ SRMR ≤ 0.10	0.05	Acceptable fit
GFI	0.90 ≤ GFI ≤ 1.00	0.85 < GFI < 0.90	0.93	Good fit
CFI	0.90 ≤ CFI ≤ 1.00	0.85 < CFI < 0.90	0.93	Good fit
IFI	0.90 ≤ IFI ≤ 1.00	0.85 < IFI < 0.90	0.94	Good fit

The initial test of the measurement model indicated a suboptimal fit to the data. Based on an examination of modification indices and theoretical considerations, the model was refined by allowing two pairs of error terms to covary. Specifically, covariances were specified between the error terms of ‘Obsession’ and ‘Disengagement’, and between ‘Lack of Control’ and ‘Escapism’. These adjustments are theoretically justifiable as these subscale pairs may share unmodeled method variance.

An examination of the goodness-of-fit indices for the measurement model revealed that χ² = 174.49, df = 49, χ²/df = 3.56, RMSEA = 0.079, SRMR = 0.05, GFI = 0.93, CFI = 0.93, and IFI = 0.94. These results indicate that the measurement model has acceptable fit values. The standardized path coefficients (p < .05) for the measurement model are presented in Fig. 2.

Fig. 2.

Standardized Path Coefficients for the Measurement Model (p < .05)

The results of the test showed that the factor loadings for the variables in the measurement model ranged from 0.52 to 0.93, and all paths were significant at the p < .05 level. Furthermore, a significant relationship was found between all latent variables in the model. The parameter estimates for the final form of the measurement model are presented in Table 4.

Table 4.

Parameter values for the measurement model

Latent Variables / Observed Variables	Standardized Loadings	t-value	R 2
Self-Efficacy
Initiative	0.71	14.06*	0.51
Effort	0.89	17.51*	0.79
Persistence	0.53	10.40*	0.28
PIU
Lack of Control	0.77	16.90*	0.59
Escapism	0.60	12.02*	0.36
Disengagement	0.65	13.99*	0.43
Obsession	0.52	10.48*	0.27
Health and Interpersonal Trouble	0.68	14.62*	0.46
General Phubbing
Nomophobia	0.80	18.86*	0.63
Interpersonal Conflict	0.74	16.84*	0.54
Self-Isolation	0.93	24.13*	0.87
Problem Acknowledgment	0.56	11.99*	0.32

Note: R² represents the proportion of variance in the observed variables; ^* p < .05

Table 4 shows that the amount of variance explained by the latent variables for the observed variables in the measurement model ranged from 0.27 to 0.87. Additionally, an examination of the relationships between variables revealed correlation coefficients ranging from − 0.39 to 0.84, with no values exceeding 0.85, which could lead to multicollinearity problems [28]. These results from the measurement model indicate that the necessary conditions for structural model analyses have been met.

Examining the mediating role of phubbing in the relationship between self-efficacy and problematic Instagram use

The mediating role of phubbing in the relationship between self-efficacy and PIU was examined using structural equation analysis. After confirming the measurement model based on the data obtained from the research, structural model analyses were conducted. The fit indices of the structural model were examined to determine how well the model fit the obtained data.

When the model fit indices in Table 5 were examined, it was observed that the constructed structural model had an acceptable fit with the obtained data (χ² = 174.49, df = 49, χ²/df = 3.56, RMSEA = 0.079, SRMR = 0.05, GFI = 0.93, CFI = 0.93, and IFI = 0.94). The tested structural model is presented in Fig. 3.

Table 5.

Fit indices for the structural model

Goodness-of-fit measures	Good fit	Acceptable fit	Model Value	Comment
χ²			174.49
Sd			49
χ²/ df	0 ≤ χ²/df ≤ 2	2 < χ²/df ≤ 0.5	3.56	Acceptable fit
RMSEA	0.00 ≤ RMSEA < 0.05	0.05 ≤ RMSEA ≤ 0.10	0.079	Acceptable fit
SRMR	0.00 ≤ SRMR < 05	0.05 ≤ SRMR ≤ 0.10	0.05	Acceptable fit
GFI	0.90 ≤ GFI ≤ 1.00	0.85 < GFI < 0.90	0.93	Good fit
CFI	0.90 ≤ CFI ≤ 1.00	0.85 < CFI < 0.90	0.93	Good fit
IFI	0.90 ≤ IFI ≤ 1.00	0.85 < IFI < 0.90	0.94	Good fit

Fig. 3.

Standardized Path Coefficients in the Structural Model (*p < .05)

When the structural model was examined, all specified paths were found to be statistically significant. The results of the structural equation model analysis are presented in Fig. 3. The analysis showed a significant direct association between self-efficacy and PIU (β = –0.07, p < .05). To test the significance of the mediating role of phubbing, a bootstrapping procedure with 5000 resamples was performed. The results revealed a significant indirect association of self-efficacy with PIU through phubbing (b = − 0.35, 95% BCa [-0.38, − 0.15]), as the confidence interval does not contain zero. The presence of a significant indirect association confirms that phubbing plays a mediating role in the relationship between self-efficacy and PIU.

Discussion and conclusion

The findings of this study contribute to the growing body of literature on problematic social media use by elucidating the interplay between self-efficacy, phubbing, and PIU, particularly within a sample of Turkish university students. This study offers novel insights by demonstrating empirically that phubbing acts as a significant mediator in the relationship between self-efficacy and PIU. While self-efficacy has been established as a protective factor against various negative outcomes, and phubbing is recognized as a detrimental social behaviour, our research uniquely positions phubbing as a behavioural mechanism through which the protective influence of self-efficacy on PIU might operate. This is a crucial finding, as it suggests that interventions aiming to reduce PIU might benefit from not only enhancing self-efficacy but also directly addressing phubbing behaviours. The results indicate that individuals reporting higher self-efficacy also tend to report being less prone to phubbing behaviour and, consequently, lower levels of PIU. This aligns with prior research suggesting an association between self-efficacy, behavioural regulation, and susceptibility to addictive-like behaviours [8]. Furthermore, the observed mediating role of phubbing highlights the potential importance of considering both psychological traits (e.g., self-efficacy) and behavioural factors (e.g., phubbing) in interventions aimed at addressing PIU.

Phubbing was found to be a significant mediator in the relationship between self-efficacy and PIU. This is consistent with the work of Chotpitayasunondh and Douglas [12], who identified phubbing as a disruptive behaviour associated with social disconnection. This finding underscores the potential need for strategies that promote mindful technology use and encourage face-to-face interactions, as suggested by Przybylski and Weinstein [39]. For instance, campaigns encouraging balanced technology use could be considered to address adverse associations of phubbing and PIU.

However, it is important to note that the cross-sectional design of this study limits causal inferences. While the findings suggest significant associations between self-efficacy, phubbing, and PIU, they should be interpreted as correlational rather than causal. Future longitudinal studies are needed to explore the temporal dynamics of these relationships and provide stronger evidence for causality.

The study also has limitations that warrant consideration. First, the sample consisted exclusively of university students, which may limit the generalizability of the findings to other age groups or populations with different socio-demographic characteristics. For example, Roberts et al. [40] found that social media addiction patterns vary across age groups, with younger individuals being more susceptible to addictive behaviours due to higher engagement with digital platforms. Second, this study focused solely on the mediating role of phubbing. Future research should investigate additional variables, such as social support, emotional regulation, and personality traits, which may also influence the relationship between self-efficacy and PIU [41, 39]. Third, as our data were collected from a single source via self-report questionnaires at one point in time, common method variance (CMV) may be a potential concern. Although we took procedural steps to mitigate this, such as ensuring anonymity, the possibility of its influence on the observed relationships cannot be entirely ruled out.

Despite these limitations, the findings have potential practical implications. Interventions aimed at supporting self-efficacy could be considered as strategies for addressing phubbing behaviour and PIU. For instance, workshops and training programs designed to foster self-efficacy might support students in regulating their digital behaviours more effectively [7]. Additionally, universities and educational institutions should consider organizing awareness campaigns to highlight the negative consequences of phubbing and excessive social media use. Peer-led initiatives and seminars could support students in developing healthier digital habits and encourage a more balanced approach to technology use.

In conclusion, this study provides valuable insights into the psychological and behavioural factors associated with PIU. By integrating these findings with existing research, future interventions might adopt a more holistic approach to addressing the complex dynamics associated with problematic social media use. Recommendations for future research include conducting longitudinal studies, exploring cross-cultural differences, and investigating the roles of additional variables such as social support and emotional regulation.

The primary novelty of this study lies in its focused exploration of phubbing as a specific mediator in the relationship between self-efficacy and Problematic Instagram Use (PIU). While previous research has examined these variables independently or explored other mediators/moderators, our study is among the first, to our knowledge, to empirically test and confirm this mediational pathway in a sample of university students, especially within the Turkish context. This integration offers a more nuanced understanding by suggesting how self-efficacy might relate to PIU – that is, partly through its association with lower engagement in phubbing. Furthermore, the study contributes by examining these contemporary digital phenomena (phubbing and PIU) in conjunction with a foundational psychological construct (self-efficacy), providing a model that can inform both theoretical development and practical intervention strategies for a population highly vulnerable to problematic social media use. The focus on a non-Western sample (Turkish university students) also adds to the diversity of research in this field, which is often dominated by studies from Western contexts. Additionally, the focus on university students provides valuable information about a demographic particularly susceptible to social media addiction. The use of structural equation modelling (SEM) supports the robustness of the analysis, allowing for a detailed examination of both direct and indirect associations. By adopting a correlational framework, this study aims to avoid overstating causal relationships and instead provides a realistic interpretation of the findings. These contributions not only advance theoretical understanding but also offer potential considerations for intervention strategies aimed at addressing PIU.

Recommendations

Based on the findings of this study, several recommendations can be made for future research and practical applications:

1. Future studies should aim to include diverse age groups and socio-demographic populations to enhance the generalizability of the findings. While this study focused on university students, exploring the relationships between self-efficacy, phubbing, and PIU in other groups, such as adolescents or working professionals, could provide a more comprehensive understanding of these dynamics.

2. This study examined the mediating role of phubbing, but other variables, such as social support, family dynamics, emotional regulation, and personality traits, could also play significant roles in the relationship between self-efficacy and PIU. Future research should investigate these factors to gain a more nuanced perspective. While this study focused on phubbing as a mediator, future research should investigate a broader range of potential mediating and moderating variables to gain a more comprehensive understanding of the relationship between self-efficacy and PIU. For instance, variables such as social support, family dynamics, emotional regulation strategies, specific personality traits (e.g., neuroticism, conscientiousness), fear of missing out (FoMO), or even different dimensions of phubbing (e.g., partner phubbing vs. general social phubbing) could be explored as additional mediators or as moderators that might alter the strength or direction of the observed associations.

3. Conducting longitudinal studies could help uncover how the relationships between self-efficacy, phubbing, and PIU evolve over time. This approach would provide deeper insights into how these variables interact and change, offering a stronger foundation for intervention strategies.

4. To establish the directionality of the observed relationships and to move beyond correlational inferences, future research should employ longitudinal designs. Tracking changes in self-efficacy, phubbing, and PIU over time would provide stronger evidence for the proposed mediational model. Furthermore, intervention studies aimed at enhancing self-efficacy or reducing phubbing behaviours could assess the subsequent impact on PIU, offering valuable insights for prevention and treatment strategies.

5. The findings suggest that higher self-efficacy is associated with lower reported phubbing behaviour and lower levels of PIU. Therefore, it is recommended to consider designing and implementing programs that focus on fostering self-efficacy through workshops, training sessions, or digital tools. These programs could also include modules on mindful technology use and the importance of face-to-face interactions.

6. Universities and educational institutions should consider organizing awareness campaigns to highlight the negative associations of phubbing and excessive social media use. Workshops, seminars, and peer-led initiatives could support students in developing healthier digital habits and encourage a more balanced approach to technology use.

7. Educational institutions and policymakers could collaborate to create guidelines or policies that encourage healthy technology use. For example, integrating digital literacy and mental health education into curricula could support students in managing their social media use more effectively.

8. Developers of social media platforms could explore features that encourage users to monitor and limit their screen time. For instance, implementing reminders for breaks or providing insights into usage patterns could support users in becoming more aware of their habits and taking proactive steps to address problematic use.

9. Conducting cross-cultural research could reveal how cultural differences may be related to the relationships between self-efficacy, phubbing, and PIU. Such studies could provide valuable insights for tailoring interventions to specific cultural contexts.

By considering these recommendations, future research and practical initiatives can build on the findings of this study to develop more effective strategies for addressing PIU and its associated negative aspects.

Author contributions

T.P.A. collected and analysed the data. E.A. conceptualized the study and designed the methodology. Ş.Ç. contributed to writing and editing the manuscript. All authors read and approved the final manuscript.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

The data that support the findings of this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki and the Data Protection Regulation (EU) 2016/679 of the European Parliament. Ethical approval for this study was granted by the Ethics Committee for Educational Research of the Ordu University (Türkiye) with number 2023/153. We obtained informed consent from all participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.