Contribution of Social Media Addiction on Intention to Buy in Social Media Sites

Abstract

The aim of this study is to assess whether social media addiction contributes to the intention to buy; it is based on the model of Hajli (2014) that assesses the relationships between the constructs of social media use, trust, perceived usefulness, and intention to buy in social media sites. To this end, a confirmatory factor analysis was carried out to evaluate whether the Hajli model applied to this sample, as well as multigroup CFA to measure invariance across gender and across following influencers or not. Finally, the path analysis evaluates the intersection of social media addiction with the Hajli model (2014). The results confirmed the Hajli model as well as the inclusion in the model of social media addiction as a variable that contributes to purchase intention on social media. Configural, metric, and scalar invariance were found across genders and across the following influencers or not. Also, the values found for internal consistency and composite reliability, convergent reliability, and discriminant reliability were within the reference values.

1. Introduction

This study posits that customers’ social activities on social networking sites influence purchasing decisions [1]. However, this can be done at the expense of addictive behavior regarding the use of social network sites (SNSs). Two types of social commerce have been identified by Huang and Benyoucef [2]: The first one takes place on sites dedicated to e-commerce, for example, Amazon. However, this type of site does not allow customers to tag others, send private messages, or initiate voting activities. The second one is based on social networking sites, integrating e-commerce features, and allowing customer connections through value creation through collaborative activities (for example, information sharing and content generation). The context of this study is the second type of social commerce.

In 2014, Hajli demonstrated the role of social media in the development of e-commerce into social commerce: “social media facilitate the social interaction of consumers, leading to increased trust and intention to buy; trust has a significant direct effect on intention to buy; the perceived usefulness of a site is also identified as a contributory factor” (p. 387). In fact, social media tools led to social commerce; social media strengthens interactions within online communities [3]. Also, social media influencers are largely used as a marketing strategy because they successfully attain trust from followers as the “influencers have sizable groups of followers and serve as experts” ([4], p. 223). Moreover, Word of Mouth (WOM) and Electronic Word of Mouth (eWOM) play a role in improving businesses and can extend their influence beyond the business realm, encompassing areas such as news, political events, scientific discoveries, etc. [5]. In turn, trust is one of the most important prerequisites for e-commerce success [6], in the social media context [7]; indeed, social media marketing directly encourages purchase intention [8]. Applications like Yelp, designed for the evaluation of commercial establishments, also contribute to enhancing business, although [9] emphasize the negative reviews on Yelp. At last, perceived usefulness impacts on intention to purchase [10].

However, the use of social networks, when it is not just for professional or academic purposes, can be problematic [11] and even become addictive [12]. Problematic social media use is a non-substance-related disorder with detrimental effects because of preoccupation and compulsion to engage in social media platforms despite negative consequences [13]. Social media addiction is a maladaptive psychological dependency on SNSs, being that behavioral addiction symptoms occur [14]. Griffiths [15] defined addictive behavior as “any behavior that features what he believes are the six core components of addiction (i.e., salience, mood modification, tolerance, withdrawal symptoms, conflict, and relapse)” ([16], p. 121). A minority of individuals use social networking sites as their most important activity, leading to a preoccupation with SNS use (salience); these activities induce mood alterations, pleasurable feelings, or a numbing effect (mood modification); increased time and energy are spent engaging with SNSs to achieve the same feelings and state of mind that happened in the initial phases (tolerance); if SNS use is interrupted, addicted individuals will feel negative psychological and physiological symptoms (withdrawal), often leading to a reinstatement of their SNS use (relapse) [17].

In addition to psychological symptoms, social media addiction may give rise to physical symptoms, such as eye strain, sleep disturbances, and increased sedentary behavior with potential health implications [18]. Behavioral symptoms, including reduced productivity, neglect of responsibilities, compulsive checking of social media accounts, and disruptions in face-to-face interactions, can also manifest [19]. Social symptoms may involve withdrawing from offline relationships, heightened feelings of loneliness, challenges in maintaining real-life connections, and potential impacts on academic or professional performance [20].

Time spent on social media predicts social media addiction [21,22]. In turn, users who spend more time on social media expose themselves to brand-/marketing-related information on SNS, resulting in enhanced purchase intentions [23]. Thus, people addicted to social networks spend more time on them and are therefore more exposed to products and brands that use social networks for their marketing, becoming more likely to buy or consume these same products.

The aim of this study is to assess if social media addiction contributes to the intention to buy, based on the model of [24] that assesses the relationships between the constructs of social media use, trust, perceived usefulness, and intention to buy in SNSs. Thus, this research postulates the following four hypotheses: ref. [25] studied the influence of different types of social media use on social media addiction and found that entertainment use is more likely to lead to social media addiction. Ref. [26] found that trust in the Internet and trust in firms significantly influence consumers’ trust and ultimately their intention to engage in social commerce. Also, ref. [27] found that the social network characteristic (informativeness) affects trust and trust has an effect on the purchase intention. Izuagbe et al. [28] found that the social influence process of subjective norms and images were determinants of the perceived usefulness of social media. We expect that (H1) social media use positively affects social media addiction (a) [25], trust (b) [26,27], and perceived usefulness (c) [28].

Ventre and Kolbe [29] found that the perceived usefulness of online reviews influences trust and online purchase intention. According to Fang et al. [30], trust can be created through online reviews. In addition, according to Mou et al. [31], perceived usefulness and trust are important in the consumer acceptance of online services because consumers’ experiences modify perceptions of usefulness. Hanaysha [32] found that trust mediates the association between two social media marketing features (interactivity and informativeness) and consumer purchase intention. We expect that (H2) trust affects positively perceived usefulness (d) [29,30,31] and intention to buy (e) [32].

People’s reliance on social media gives rise to social media dependency [33]. Social media addiction affects purchase intention [34]. We expect that (H3) social media addiction positively affects the intention to buy (f) [33,34]. Perceived usefulness is an important factor in customers’ intention to purchase via social media use [35,36]. We expect that (H4) perceived usefulness positively affects the intention to buy [35,36].

2. Methods

2.1. Procedures

All the procedures took into consideration the most up-to-date version of the Declaration of Helsinki (2013) (and all the ethical principles underlying it). This study was approved by the Scientific Council of the Portuguese Catholic University. The research protocol was applied after the participants signed the informed consent including the purpose of the study, guaranteeing anonymity, confidentiality, and that all information collected would only be used to pursue the research objectives.

A research protocol was developed, including a sociodemographic questionnaire, the Hajli [24] instrument to measure the relationships between the constructs of social media use, trust, perceived usefulness, and intention to buy in SNSs, and the Social Media Addiction Scale (SMAS, [37]; Portuguese version by Leite et al. [38]). The protocol was made available online through Google Forms and was posted on a page on a social network created specifically for this purpose. Its dissemination was made mainly through personal contacts, using social networks and other digital means, and obtaining a sample through the snowball effect, which is non-probabilistic for convenience. Data collection took place during January 2023. The data obtained were processed using the Statistical Package for the Social Sciences (SPSS, version 28) and AMOS (version 28).

2.2. Sample

The sample consists of 639 participants, about 65% female (n = 418). The age ranged from 16 to 76 years, and the approximate average age, 26 years (M = 25.72, SD = 11.19), was more than the median age (21 years) of the sample. The majority (51%; n = 328) attend higher education and are professionally active (student or employed) (78%; n = 500).

This sample preferably uses the social media Instagram (64%; n = 409), followed by Facebook (27%; n = 174), followed by Twitter (8%; n = 48), and others (1%; n = 8). Most of the sample participants used social media on a daily basis (61%; n = 390) and followed influencers on social media (70%; n = 445).

2.3. Instruments

The sociodemographic questionnaire included issues related to gender, age, and education. This questionnaire also included questions related to social media use on a daily basis, following influencers in social media, and the preferred one (Instagram, Facebook, Twitter, and others).

2.3.1. Instrument to Assess the Relationships between the Constructs of Social Media Use, Trust, Perceived Usefulness, and Intention to Buy in SNSs (TPSS)

Hajli [24] developed a research model to examine the relationships between the constructs of social media use, trust, perceived usefulness, and intention to buy in SNSs. This model was tested based on a questionnaire that Hajli [24] also developed, using a 5-point Likert scale from 1 = strongly disagree to 5 = strongly agree. The items were adopted from previous research. Social media use assesses whether the person uses social networks and how much time they spend on them. Trust, based on several previous studies, measures trust in SNSs and peers in these networks; social interaction measures the activity of individuals through online communities, forums, ratings, reviews, and recommendations [24]; perceived usefulness measures a person’s level of confidence in that using a particular technology he will improve his performance (Van der Heijden, 2003) [39]; intention to buy measures the user’s willingness to purchase on SNSs and their intention to buy through SNSs. Hajli’s model (2014) presented very good psychometric properties, namely, a Cronbach’s alpha above 0.70 for each dimension: social media use (α = 0.87), trust (α = 0.81), perceived usefulness (α = 0.88) and intention to buy in SNSs (α = 0.76).

2.3.2. Social Media Addiction

The Social Media Addiction Scale (SMAS, [37]; Portuguese Version by Leite et al. [38]) was adapted from the Young’s Internet Addiction Scale [40]. This instrument evaluates the symptoms of addiction to social media through three factors, and fourteen items, answered on a 5-point Likert scale ranging from “1—strongly disagree” to “5—strongly agree”. Factor 1 includes the deterioration of school performance, driving, not meeting friends, and thinking about social media when not using them (items 1, 2, 3, and 4) (α = 0.70). Factor 2 involves social media overuse, neglecting schoolwork, feeling irritable, and lacking sleep due to social media usage (items 5, 6, 7, and 8) (α = 0.70). Factor 3 has two items, one dealing with boredom and the other with the need to use social media (items 9 and 10) (α = 0.70) [37]. Al-Menayes [37] excluded Young’s items (6, 8, 9, and 14) because they presented low loadings (under 0.50). High values are associated with high addiction to social media.

The Portuguese version of this scale [38] includes ten items distributed by two factors: factor 1 (items 1, 2, 6, 11, 13, and 14) named compulsive feelings (α = 0.76) and factor 2 (items 3, 4, 8, and 9) named social consequences (α = 0.73) and the total scale presented a Cronbach’s alpha of 0.81. The scale instruction and response modalities were maintained in the Portuguese version.

2.4. Data Analysis

Descriptive statistics were used to describe the study participants’ characteristics. Continuous data were reported as means and standard deviations and categorical data were reported as frequencies and percentages.

Confirmatory factor analysis (CFA) assessed the fourth-dimensional factor structure of Hajli’s model [24]. Several indices were used to assess the model fit, including the comparative fit index (CFI) and the Tucker–Lewis index (TLI) > 0.9; root mean square error of approximation (RMSEA) < 0.06; standardized root mean square residual (SRMR) < 0.06; and non-significant chi-square [41,42]. Multigroup CFA analyses were conducted on Hajli’s model across men and women, and across those who follow influencers and those who do not, to assess the invariance between groups; the multigroup analysis allows testing if pre-defined groups have significant differences in their parameter estimates. The configural invariance test allows the examination of whether the overall previous factor structure fits well for both genders, allowing all factor loadings and item intercepts to vary freely for each group. Metric invariance examines whether the factor loadings are equivalent across the groups. Scalar invariance assesses whether the item intercepts are equivalent across groups [43].

Internal consistency (Cronbach’s α > 0.7 is acceptable) [44], composite reliability (CR; values higher than 0.6 indicate acceptable reliability), average variance extracted (AVE; values > 0.5 indicate satisfactory convergent reliability), and square root AVE (higher than inter-construct correlations indicate satisfactory discriminant reliability) [45] were assessed.

A path analysis was carried out to analyze and confirm the model present in the literature. Path analysis is a form of multiple regression statistical analysis used to evaluate causal models, examining the relationships between a dependent variable and two or more independent variables. Model fit was assessed using the following criteria: a chi-square/df ≤ 2, a p-value > 0.05, a comparative fit index ≥ 0.95, a Tucker–Lewis Index ≥ 0.95, and a root mean square error approximation of <0.06.

3. Results

3.1. Confirmatory Factorial Analysis (CFA)

To confirm the instrument proposed, a CFA was carried out (Figure 1), presenting a good model adjustment (χ²(59) = 2.851; p < 0.001; CFI = 0.973; TLI = 0.964; RMSEA = 0.054; PCLOSE = 0.241: SRMR = 0.069).

3.2. Measuring Invariance through Multigroup CFA

Multigroup CFA was conducted on the TPSS between men and women (

Table 1. Multi-group confirmatory factor analysis (MGCFA) across gender.

Invariance	RMSEA	SRMR	CFI	TLI	Model Compare	ΔRMSEA	ΔSRMR	ΔCFI	ΔTLI	Decision
M1: Configural	0.042 (0.034–0.049)	0.075	0.968	0.958
M2: Metric	0.040 (0.033–0.047)	0.079	0.968	0.961	M1	−0.002	0.004	0.000	0.003	Accept Δ < 0.010
M3: Scalar	0.040 (0.033–0.047)	0.093	0.967	0.960	M2	0.000	0.014	−0.001	−0.001	Accept Δ < 0.030

Note: RMSEA = root mean square error of approximation; SRMR = standardized root mean squared residual; CFI = comparative fit index; TLI = Tucker–Lewis index.

). The model found presented a very good fit to the data (χ²(118) = 2.092; p < 0.001; CFI = 0.968; TLI = 0.958; RMSEA = 0.042; PCLOSE = 0.973; SRMR = 0.075). Models for men and women are presented in Figure 2 and Figure 3, respectively. The multigroup test for configural invariance showed support for accepting the hypothesis of configural equivalence (all differences between fit indexes were acceptable) among factor structures; thus, participants from different groups conceptualized Hajli’s model [24] in the same way. The comparison between configural and metric models indicates that metric invariance was achieved because factor loadings were stable across genders. Also, the comparison between the metric and scalar models suggested that scalar invariance was achieved (

Table 2. Multi-group confirmatory factor analysis (MGCFA) across followers (or not) of influencers in social media.

Invariance	RMSEA	SRMR	CFI	TLI	Model Compare	ΔRMSEA	ΔSRMR	ΔCFI	ΔTLI	Decision
M1: Configural	0.041 (0.033–0.048)	0.089	0.968	0.958
M2: Metric	0.039 (0.032–0.046)	0.090	0.969	0.962	M1	0.002	−0.001	−0.001	−0.004	Accept Δ < 0.010
M3: Scalar	0.044 (0.037–0.050)	0.086	0.956	0.951	M2	−0.005	0.004	0.013	0.011	Accept Δ < 0.030

Note: RMSEA = root mean square error of approximation; SRMR = standardized root mean squared residual; CFI = comparative fit index; TLI = Tucker–Lewis index.

). Differences in CFI, TLI, and RMSEA values were all within the cut-off range suggested by Rutkowski and Svetina [46]. There are statistically significant mean differences between men and women concerning the total scale (t(634) = −2.053; p = 0.040; d = 0.571) and the social media use subscale (t(386, 621) = −2.707; p = 0.007; d = 0.765), with women scoring significantly higher than men in both dimensions.

Multigroup CFA was also conducted on the TPSS concerning participants without following influencers and participants that follow influencers on social media. The model found presented a very good fit to the data (χ²(90) = 2.050; p < 0.001; CFI = 0.968; TLI = 0.958; RMSEA = 0.041; PCLOSE = 0.984; SRMR = 0.089). The multigroup test for configural invariance showed support for accepting the hypothesis of configural equivalence (all differences between fit indexes were acceptable) among factor structures; thus, participants from different groups conceptualized the construct TPSS in the same way. The comparison between configural and metric models indicates that metric invariance was achieved because factor loadings were stable across participants who follow influencers and other participants who do not follow them. Also, the comparison between the metric and scalar models suggested that scalar invariance was achieved (

Table 3. Correlations, AVE, AVE square roots, CR, mean, standard deviation, and Cronbach’s alpha of TPPS total and subscales.

	1	2	3	4	5	AVE	CR	M	SD	α
1 TPPS Total	0.589					0.342	0.868	3.20	0.57	0.83
2 Trust	0.688 **	0.846				0.716	0.909	3.18	0.72	0.86
3 Purchase Intention	0.724 **	0.391 **	0.820			0.672	0.804	2.85	0.86	0.51
4 Perceived Usefulness	0.745 **	0.298 **	0.565 **	0.868		0.753	0.901	2.95	0.94	0.84
5 Social Media	0.682 **	0.230 **	0.305 **	0.286 **	0.767	0.588	0.850	3.57	0.77	0.76

Note: ** p < 0.001 for Pearson correlation; AVE = average variance extracted; bold = AVE square roots; CR.

).

There are statistically significant mean differences between participants that follow influencers and those who do not follow them concerning the total scale (t(637) = −6.535; p < 0.001; d = 0.554), trust subscale (t(637) = −3.632; p < 0.001; d = 0.716), purchase intention subscale (t(637) = −4.477; p < 0.001; d = 0.850), perceived usefulness subscale (t(637) = −5.260; p < 0.001; d = 0.922), and the social media use subscale (t(325, 190) = −4.546; p < 0.001; d = 0.756), with participants that follow influencers scoring significantly higher than those who do not, in all dimensions.

3.3. Convergent and Divergent Validity

Reliability was assessed by Cronbach’s alpha presenting values above those recommended, except for purchase intention. Convergent validity was assessed by composite reliability (CR) (whose values were above the reference ones) and average variance extracted (AVE) values (whose values were over 0.500 except for Hilaj’s model total). Discriminant validity was assessed by the square roots of the AVE values (Table 3); all these values were higher than the correlation values of each construct except for Hijal’s model total.

3.4. Associations between Hajli’s Model and Social Media Addiction

There are positive and significant correlations between almost all dimensions of Hajli’s model and the SMA, except for the SMA social consequences that do not correlate with trust and social media use. Furthermore, the correlations are weak (between r = 0.087 and r = 0.273) (

Table 4. Correlations between TPPS and subscales and SMA and subscales.

	Total TPPS	TPPS Trust	TPPS Purchase Intention	TPPS Perceived Usefulness	TPPS Social Media Use
Total SMA	0.223 **	0.129 **	0.242 **	0.127 **	0.160 **
SMA Compulsive Feelings	0.273 **	0.201 **	0.261 **	0.123 **	0.203 **
SMA Social Consequences	0.087 *	−0.011	0.143 **	0.094 *	0.052

Note: ** p < 0.001 for Pearson correlation; * p < 0.01 for Pearson correlation.

).

3.5. Path Analysis

A path analysis was carried out to analyze and confirm the model presented in the literature by Hajli [24]. The model of this author was used, but this study added a variable to the model: social media addiction. This variable was added to the model as a variable that could contribute to purchase intention. The model found fits the data well (χ²(3) = 3.035; p < 0.001; CFI = 0.987; TLI = 0.935; RMSEA = 0.056; PCLOSE = 0.222). All our hypotheses were confirmed as can be seen in Figure 4.

4. Discussion

The aim of this study is to assess whether social media addiction contributes to the intention to buy; it is based on the model of Hajli [24] that assesses the relationships between the constructs of social media use, trust, perceived usefulness, and intention to buy in social media sites. First of all, the model proposed by Hajli [24] was confirmed. Like the author, our results showed that social media use facilitates the social interaction of consumers, leading to increased trust and intention to buy; trust has a significant effect on the intention to buy, and the perceived usefulness of a site is also a contributory factor for the intention to buy. Hajli [24] stated that the theoretical implication of his model was integrating a technology acceptance model (TAM) with “trust and social media use to develop a model for adoption in social commerce” (p. 401).

In our study, as in the Hajli study [24], convergent validity was ensured by assessing discriminant and convergent validity. The values found for internal consistency and composite reliability, convergent reliability, and discriminant reliability were within the reference values. In our study, these values fully corroborate those found by Hajli [24].

Configural, metric, and scalar invariance on the TPSS were found concerning gender and following influencers or not, providing strong evidence that TPSS operates similarly in different genders and in consumers that do or do not follow influencers in social media. Our results showed that women use social media more than men. This is in line with Noori et al. [47] who also found that female participants used social media more frequently than male. Also, Hewei and Youngsook [48] found that, compared with men, women present a higher average value on social media interaction and continuous purchase intention. Su et al. [49] studied whether males and females differ in specific patterns and types of internet usage and related internet addiction and found that higher levels of internet gaming disorder (IGD) exist in males and higher levels of social media addiction exist in females.

Our results also showed that participants who follow influencers score significantly higher than those who do not in social media use, trust, perceived usefulness, and intention to buy. According to Morton [50], the motivations to follow social media influencers are information, inspiration, communication, entertainment, and surveillance. Also, Croes and Bartels [51] found six other motivations to follow influencers: information seeking and information sharing, companionship, cool and new trends, relaxing entertainment, and boredom/habitual pass time. Also, the portrayal of death and grief is becoming more prevalent in digital media; concurrently, it is reshaping the dynamics of interaction among those mourning their loss [52]. Attributes of influencers and follower participation foster attachment development and attachment causes problematic behaviors such as excessive use [53], which, as we have already seen, exposes social media users to more products leading to a greater purchase intention (presupposing more social media use, more trust, and more perceived usefulness).

However, the most important result of this study concerns the fact that addiction to social media positively affects purchase intention. According to the theory of media dependency [54], the more people rely on the media to meet their needs, the more important the media will be in their lives and have more impact on them [55]. In addition, according to the media dependency theory applied to the digital context, “the greater the consumer’s connection with social media, the greater the likelihood that they will be involved in purchase-related activities” ([56], p. 368). Farzin et al. [57] stated that social media-associated features (reduced perceived risk and consumer trust) influence purchase intention in social media; also, addiction, usability, dependency, and involvement affect the level of trust of consumers and their perceived risk in social media sites.

Generally, researchers are optimistic concerning the relationship between social media use and purchase behavior because of the value of the investment in social media advertisements [58]; however, Keles et al. [59] showed that social media use is classified into four domains, investment, time spent, activity, and addiction, and concluded that depression, anxiety, and psychological distress are associated with all four social domains. Sun and Zhang [14] reviewed 25 theories/models applied in studies of social media addiction and defined “addictive social media use as being overly concerned about social media, strongly motivated and having been devoting a great amount of time and energy to use social media, to the degree that an individual’s social activities, interpersonal relationships, studies/jobs, and/or health and well-being are impaired” (p. 2). The question that arises is whether it is legitimate for companies to continue to optimize the use of online social networks (and, obviously, their addictive capacity) to increase sales and, ultimately, profits. This is an ethical issue and not just a clinical or economic one. According to Bhargava and Velasquez [60], “addicting users to social media is impermissible because it unjustifiably harms users in a way that is both demeaning and objectionably exploitative. Importantly, the attention-economy business model of social media companies strongly incentivizes them to perpetrate this wrongdoing” (p. 321).

The present study, although it offers significant contributions to research, presents some limitations that should be discussed. Despite the fact that the sample size was substantial, it is crucial to consider the specific characteristics of sociodemographic elements. The presence of elements such as specific age groups, gender distribution, and educational backgrounds in the sample may limit the generalizability of the results to a broader population. Therefore, it is imperative to exercise caution in interpreting and generalizing the obtained results. In future investigations, it would be pertinent to include participants from diverse demographic groups, encompassing different age groups, education levels, ethnicities, socioeconomic status, and other sociodemographic relevant variables. This approach would enable a more comprehensive analysis of the effects of social sites and addiction in various segments of the population.

Another important consideration relates to the influence of culture on the study’s results. The results of this study stem from a specific context, which may hinder their direct translation to diverse populations or different cultural environments. Therefore, a promising area for future research is cross-cultural comparison. Culture plays a crucial role in attitudes towards social sites and consumption. Consequently, conducting comparative studies in different cultural contexts can enhance the understanding of variations in the relationship between social media addiction and purchase intention.

Furthermore, although the self-report instruments used in data collection had appropriate psychometric properties for the research objectives, it is essential to acknowledge the inherent subjectivity of these instruments, which can introduce response biases. Additionally, participants may have encountered challenges in responding to the questionnaire as the investigator was not present to clarify any potential doubts, which could have impacted the accuracy of the collected data.

The cross-sectional design of this study represents another limitation, as it prevents the determination of causality relationships among the variables under study. Hence, the need for and relevance of conducting longitudinal studies to identify the direction of causality in the relationships between social media addiction, trust, perceived utility, and purchase intention continues to be a limitation. Moreover, monitoring the dynamics of addiction over time would provide a deeper understanding of how social media addiction influences purchase intention in a long-term context.

Other fundamental aspects to consider are the dynamic nature of the concept of social media addiction, as it is continually evolving in research, as well as the constant modification of social interactions due to platform updates or external events. In this regard, studies investigating how addiction patterns evolve in response to these elements can provide crucial information for understanding these phenomena and their interrelationships.

Finally, it is important to highlight the ethical challenges associated with the addictive nature of social media and its potential impact on consumer behavior. This ethical dimension suggests the need for further exploration in future research, going beyond purely clinical or economic aspects. Considering these ethical challenges is essential for a comprehensive analysis of the implications of social media addiction.

5. Conclusions

This paper assessed whether social media addiction contributes to the intention to buy; it is based on Hajli’s [24] model, which explores the relationships between social media use, trust, perceived usefulness, and intention to buy. This study confirms Hajli’s model, indicating that social media use facilitates consumer social interaction, leading to increased trust and intention to buy. Trust and perceived usefulness also play significant roles in influencing the intention to buy. This study ensures convergent validity by assessing discriminant and convergent validity, with values aligning with reference standards. Invariance tests on the TPSS across gender and follower status indicate consistent operation in different gender groups and among customers following or not following influencers.

Results show that women use social media more than men, which is in line with other studies. Participants following influencers score higher in social media use, trust, perceived usefulness, and intention to buy. This study suggests that attachment development due to influencer attributes and follower participation leads to excessive use, exposing users to more products and increasing purchase intention. A crucial finding is that social media addiction positively affects purchase intention, aligning with the theory of media dependency. Researchers are optimistic about the relationship between social media use and purchase behavior, but concerns arise regarding the ethical implications of companies optimizing social media’s addictive capacity for profit.

This study acknowledges limitations related to sample characteristics, emphasizing the need for caution in generalizing results. Future research should include diverse demographic groups to provide a more comprehensive analysis of the effects of social media and addiction across different segments of the population.