Psychometric Properties of the Social Media Addiction Scale (SMAS) on Chilean University Students

Introduction: the use and abuse of social networks are harming the mental health of university students. Objective: to adapt and validate the Social Media Addiction Scale (SMAS) for the Chilean context to have a reliable instrument to measure addiction to social networks. The sample comprised 686 university students (mean age=28,04, SD=8,4), 71,1 % female, 28,4 % male, and 0,5 % other genders. Methods: confirmatory factor analysis (CFA) using the weighted least squares means and variances method (WLSMV) was used for this study. Results: reliability was Cronbach’s alpha α=0,841. The SMAS yielded two factors that explained 53,433 % of the variance. The CFA yielded very good fit indicators such as CFI=0,959, TLI=0,949, and RMSEA=0,060. Conclusions: based on the results described above, we can affirm that the SMAS is a good instrument to measure social network addiction in college students.

Local and global studies have shown that depressive, anxious, and alcohol-dependence disorders are the most frequent in the university population (Martínez-Líbano et al., 2023;Richardson et al., 2017;Tembo et al., 2017).The results of the National Health Survey ENS Chile 2009-2010 present Chile with a prevalence of depressive symptoms of 14,4 % in the population aged 15-24 years (Cristi-Montero et al., 2019; Garrido-Méndez et al., 2019).In the same 2016-2017 survey, women appear with a prevalence of 21,7 % in depression and 10,1 % in men (Nazar et al., 2021).In addition to the above, a study on mental health disorders in university students showed that high alcohol consumption and deficits in self-care and recreational behaviors negatively affect students' mental health.These are causes of depressive and anxious symptomatology between 20 % and 30 % of university students; among them, first-year students stand out with 60 % of symptomatology (Antúnez Sanhueza & Vinet, 2013).The same mental health disorders have been aggravated in recent studies in Chilean university students.In 2022, the prevalence of mental health problems in this population was high, with depressive symptoms in 63,1 % of the sample, 69,2 % with anxiety, 57 % with stress, 27,4 % with problematic alcohol consumption, and 14,9 % with inappropriate marijuana use.Some 10,1 % of the sample reported daily medication with antidepressants and/or anxiolytics (Martínez-Líbano et al., 2023).In addition to the above, the results of a study related to the presence of depressive symptomatology show higher rates in the university collective compared to the general population, highlighting, in addition, a greater presence of depressive symptomatology in women than in men (Rossi et al., 2019).Finally, in recent years it has been possible to prove that the pandemic generated serious consequences in the mental health of higher education students, with severe prevalences of depression, anxiety, stress, emotional exhaustion, and suicidal ideation (Martínez-Líbano et al., 2021, 2023; Martínez-Líbano & Yeomans Cabrera, 2021), which has been extended to the new post-pandemic global context.
Worldwide, new information and communication technologies (ICTs) and ease of access have led to an increase in the number of Internet users.In August 2022, social network users worldwide accounted for 4,62 billion, equivalent to 58,4 % of the population.In Chile, in January 2022, there were 17,85 million social network users, equal to 92,8 % of the total population (Martínez-Libano et al., 2023).The Internet and new technological media have positively and negatively affected university students.The Internet has broadened horizons and facilitated communication; however, the adverse effects of excessive use of networks are seen in both the health of the mind and body of this population: year after year, university dropout rates increase (Martínez-Líbano et al., 2022; J. Wang, 2017).In general, Internet addiction is characterized by symptoms such as obsessive thoughts about the Internet, tolerance (need to increase the "dose" to obtain the same result), withdrawal syndrome, uncontrolled urges, and inability to disconnect; sometimes, this involves compulsive web surfing, obsessive use of websites for gaming, gambling, and shopping, and impulse control problems (Metin-Orta, 2020).Because the Internet is not a substance, it is considered a behavioral addiction; this addiction involves compulsive behaviors, a lack of interest in offline activities, and, when attempts are made to stop this behavior, physical and mental consequences, especially in the youth and young adult population (D'Arienzo et al., 2019).A recent study examined the association between problematic Internet use, psychological distress, and coping strategies in college students during the pandemic, showing that problematic Internet use was associated with high levels of psychological distress, self-distraction coping strategies, and behavioral disengagement; 32,2 % of participants had problematic Internet use, and 61,3 % reported psychological distress; younger students had the highest levels of psychological distress (Yusof et al., 2022).
Similarly, problematic smartphone use has developed as an essential part of Internet addiction, described as the inability to regulate device usage time.The negative implications of problematic use in daily life put users' mental health at risk (Coskun & Muslu, 2019).The features of smartphones, the portability of the Internet, and the possibility of having customized applications transform it into a versatile and multipurpose device (Khan et al., 2018;Raj & Raman, 2017).It is due to the varied uses of smartphones that smartphone addiction presents different symptoms than internet addiction: Internet addiction can occur on different types of devices as long as there is internet, and smartphone addiction occurs with the device, whether or not it has internet (Tateno et al., 2019).Haug et al. 2015 indicated that the most relevant smartphone functions for users that were positively associated with device addiction were (1) increased smartphone usage time on a typical day, (2) reduced time spent waiting to use the smartphone first thing in the morning, and (3) reporting social networks as the most relevant personalized function of the device (Haug et al., 2015).To this, Kuss and Griffths 2017 added that social networks are highly relevant to excessive smartphone use (Kuss & Griffiths, 2017).
In 2017, it was estimated that the average daily use of social networks was 135 minutes per day, which implies an increase of 9 % over the previous year (D 'Arienzo et al., 2019).Other more current figures for the Data and Metadata.2023; 2:91 2 year 2021 show that the social network with the highest number of monthly users in the world is Facebook, with 2200 million users approximately, followed by WhatsApp, with 500 million, Twitter with 284 million, and Instagram with 259 million (Stanculescu & Griffiths, 2021).In 2023, this ranking has changed.However, Facebook is still in the lead.The social network with the most users worldwide is Facebook, with approximately 2,96 billion active users worldwide.This represents an increase of 1,65 % over the previous year.The complete ranking of the top 5 social networks with the most users in the world is as follows: Facebook (2,96 billion), YouTube (2,562 billion), WhatsApp (2 billion), Instagram (1,478 billion) and TikTok (1 billion) (Kemp, 2023).It is also estimated that, globally, at least 39 % of the population is currently an active user of social networks (Savci & Griffiths, 2021).
Social network addiction, like Internet addiction, is a behavioral phenomenon characterized by excessive preoccupation with social network use and an uncontrollable need to connect to or use social networks (Quaglieri et al., 2021).Users may develop tolerance and dependence symptoms despite experiencing adverse consequences due to excessive use (Cheng & Lau, 2022).The literature exposes that social network addiction is associated with elements such as low self-esteem ( ).On the other hand, Prakabar (2017) evaluated the effect of Facebook on the social life, health, and behavior of a group of students.It was found that most of the participants complained of headaches, back pain, mood swings, eye problems, and decreased work potential due to excessive use of computers and mobile devices to manage Facebook (Prabakar, 2017).In the study by Farooqui et al. 2013, researchers found that most undergraduate medical students reported using Facebook daily and that 40,1 % spent 3-4 hours exclusively on this activity.In addition, 39 % considered themselves shy in real life, but 60,3 % of the participants were deemed funny by their Facebook friends.Despite all the above, 67 % of the participants in this study did not consider themselves addicted to the social network Facebook (Farooqi et al., 2013).Regarding Instagram and mental well-being, a group of university students expressed that Instagram was positive for the development and maintenance of friendships (Ponnusamy et al., 2020), but at the same time, they considered that this social network had multiple implications for mental well-being due to the presentation of physical ideals, the search for social acceptance, and cyberbullying.Regarding the latter, it was found that 56,3 % of the participants witnessed or received hate through the social network (56,3 %) (Javornik et al., 2022; Popat & Tarrant, 2023).In addition to the above, most expressed that they only presented their best version on this network to show off to their peers.They also perceived an increase in Instagram "likes" associated with peer validation, popularity, and likeability (Moreton & Greenfield, 2022).
Greater problematic social media use was significantly associated with greater depression, anxiety, and insomnia but not with stress.Stress mediated the relationship between depression, anxiety, insomnia, and problematic social media use (Malaeb et al., 2021).However, another study reveals that whether this effect is beneficial or detrimental depends, at least partly, on the quality of social factors in the social networking environment (Seabrook et al., 2016).It was also found that females, college students, and students with low academic achievement had the highest levels of social network addiction (Al-Menayes, 2015).Critical indicators concerning social network addiction are increased daily time spent on social networks, increased time spent on each access, increased frequency of access, increased duration of networking before bedtime, and decreased time between awakening and first network use (Reid Chassiakos et al., 2016).Other researchers suggest the need to detect and address excessive social network use to prevent social network addiction and mental distress among youth (T.Wang et al., 2021).
Continuing with the distress factor, the results of another study related to social network addiction and romantic relationship satisfaction showed that depression, anxiety, and stress partially mediated the impact of social network addiction and romantic relationship satisfaction (Spencer et al., 2017), also finding significant associations between social network addiction and relationship satisfaction through psychological distress (Márquez Benza, 2021).In other words, a high level of social network addiction leads to decreased relationship satisfaction, partially mediated by psychological distress (Satici et al., 2021).
On the other hand, reducing the use of social networks can benefit young college students' mental health (Hou et al., 2019).Most students reported a positive change in mood, reduced anxiety, and improved sleep during and immediately after the social network detoxification period (Radtke et al., 2022).Unexpected findings showed that although Facebook is the most used social network, Instagram was the social network identified as the most difficult to stop using by respondents (Anderson & Jiang, 2018), which has led students to consider social network detoxification in the future (El-Khoury et al., 2021).
Based on the above and the severe consequences established in the mental health of young people (Martínez-Líbano et al., 2022), it is necessary to have effective and validated tools regarding social network addiction to continue developing research on the consequences in the general population, so the psychometric analysis focused on the Social Network Addiction Scale (SMAS).It was hypothesized that the SMAS would present a https://doi.org/10.56294/dm202391bifactorial structure with high concurrent validity with the DASS-21, already validated in Chile.

Design
The purpose of the present instrumental study was to determine the psychometric properties of the Social Media Addiction Scale (SMAS) in a sample of 686 university students.This research was approved by the Bioethics Committee of the Faculty of Education and Social Sciences of Universidad Andrés Bello on August 18, 2023.

Participants
The sample consisted of 686 university students with a mean age of 28,04 years and a statistical deviation of 8,4 years; 71,1 % were female, 28,4 % were male, and 0,5 % were persons who did not identify with either of the two previous sexes.Inclusion criteria consisted of being a university student 18 or older.Exclusion criteria were incomplete questionnaires.

Instruments
The Social Media Addiction Scale (SMAS): This scale consists of 14 items, where appreciation of the use of social networks and ways of using them are expressed (J. Al-Menayes, 2016).Some of the statements exposed in this scale are; "I tend to use social networks more time than expected, "I tend to find life boring without social networks," and "I tend to neglect my work or schoolwork because of my use of social networks."The reliability level (α) of the scale in this study was 0,832.

Depression, Anxiety, and Stress Scale (DASS -21):
This psychometric scale consists of 21 items, in which contextual statements are used to describe the mental state that the person has recently experienced (Antúnez & Vinet, 2012).Some of the representative statements in this questionnaire concerning recent experiences are: "I found it difficult to motivate myself to do things," "I got out of control in certain situations," "I felt that I was very nervous," and "I was worried about situations in which I could panic and make a fool of myself."The scale's reliability (α) in this study was 0,95.

Procedure
The Social Media Addiction Scale (SMAS) was translated by two professional and native translators, after which the translations were analyzed by two experienced researchers who approved the application.After this, we applied the SMAS and the Depression, Anxiety, and Stress Scale (DASS -21) to a group of higher education students.Our data collection method was carried out in April and May 2023 through a self-applicable online survey published on the Google Forms platform and disseminated through the Internet on platforms such as Email, Instagram, Facebook, and WhatsApp.
The complete sample, 686 Chilean higher education students, was divided into two halves randomly using the SPSS 29 program.Therefore, to perform the exploratory factor analysis, 343 participants were used, and for the confirmatory factor analysis, the other 343 higher education students were used.
Frequency distributions for the demographic characteristics and completion rates of the SMAS scores were performed using SPSS program version 29.The internal consistency of the SMAS was assessed using Cronbach's alpha and the inter-item correlations using Spearman's rank correlation coefficients (Xiao et al., 2016).Bartlett's sphericity test and Kaiser-Meyer-Olkin's measure of sampling adequacy were used to determine the adequacy of the factor analysis (Shrestha, 2021).
The factor structure of the SMAS was examined by performing an exploratory factor analysis by halves and a confirmatory factor analysis using SPSS AMOS.An exploratory factor analysis was performed using maximum likelihood estimation with geomin oblique rotation (Y.Chen et al., 2019).Screen plots, eigenvalues (≥1), factor loadings (>0,4), and model fit of one-and two-factor structures were found.Confirmatory factor analysis was then performed to assess model fit.To determine model fit, chi-square model fit statistics (X 2 ), comparative fit index (CFI), Tucker-Lewis index (TLI), root mean square error of approximation (RMSEA) with 90 % confidence intervals, and standardized root mean square residuals (SRMR) were found.The goodness-of-fit criteria of Kline

RESULTS
The descriptive results are shown in Table 1.

Reliability Results
Reliability was measured by Cronbach's alpha, which yielded a result of 0,832, corresponding to excellent reliability (Ismail et al., 2020).Factor Analysis of the Scale KMO tests and Bartlett's test were performed to assess the suitability of the data for a factor analysis (Napitupulu et al., 2017).KMO values range from 0 to 1, where values closer to 1 indicate a higher adequacy of the data (Watson, 2017).Generally, a KMO value higher than 0,6 or 0,7 is considered acceptable to perform a factor analysis instead (Lemos Hoyos & Londoño Arredondo, 2006).For our study, the KMO was 0,860, well above what is required.Bartlett's test evaluates the null hypothesis that the correlation matrix of the variables is an identity matrix, which would indicate no relationship between the variables (Odoi et al., 2022).For our study, Barlett's test was significant (0,000).Therefore, based on the above background, the high KMO value and the significant Bartlett's test indicate we can perform the factor analysis.Once the tests were performed, the matrix was rotated to interpret better and simplify the extracted factors.From the above, it was possible to demonstrate the existence of two factors that explained 53,433 % of the variance.After performing the rotation, two factors appeared: factor 1 is called "disconnection from reality due to the use of social media" (DRSM), and factor 2 is called "waste of time due to the use of social media" (WTSM).

Confirmatory Factor Analysis
A confirmatory factor analysis was performed with the other part of the sample to confirm the two-factor structure found in the exploratory factor analysis of the SMAS.The confirmatory factor analysis is presented in Figure 1.

Concurrent Validity
Concurrent validity was performed with the DASS-21 with good levels of correlation between the two factors yielded by the SMAS and the variables measured by the DASS-21, such as depression, anxiety, and stress (see Table N°6).From the exploratory factor analysis, it was possible to determine the existence of two factors related to addiction to social networks: factor 1, disconnection from reality due to the use of social media (DRSM), and factor 2, waste of time due to the use of social media (WTSM).
The disconnection from reality through social networks can be understood, given that sometimes people show an idealized version of their own self through social networks (Chua & Chang, 2016;Jin et al., 2019).The above could even generate unfavorable comparisons (Fardouly et al., 2015) and dissatisfaction with one's own life (Abel et al., 2016), resulting in a sense of disconnection and permanent loneliness in people who try to disconnect from reality.Although social networks allow an apparent connection with others (Carr & Hayes, 2015), the interaction that occurs on these platforms is often more superficial and lacks the richness of face-to-face interactions (Bibi et al., 2018;McCrae et al., 2017), generating in people who feel an emotional disconnection.
The second factor (time wasted using social networks) appears when some people become dependent on social networks, spending long hours on them (Yesilyurt & Solpuk Turhan, 2020) and neglecting other important activities and relationships (Han et al., 2019).This dependence can lead to feelings of isolation and disconnection from reality (Tsay et al., 2013).
Concerning the confirmatory factor analysis, it was found that this scale presents very good fits, all above 0.90, which is the standard of adequate fit for psychometric scales.Therefore, it is a good instrument that correctly measures addiction to social networks.
The validation of psychometric scales is fundamental in scientific research, as it helps us confirm the consistency and stability of the construct we wish to measure and gives us confidence that the results will be valid, thus contributing to the study of new phenomena.

CONCLUSION
The SMAS is a scale with good psychometric properties for measuring addiction to social media.It has good reliability (Cronbach's alpha (α) 0,841.The CFA yielded very good fit indicators such as CFI=0,959, TLI=0,949, and RMSEA 0,060 and presents two factors: (1) Disconnection from reality due to social media use (DRSM) and (2) Waste of time due to social media use (WTSM), which explain 53,433 % of the variance.It also presents a good psychometric fit and concurrent validity with the DASS-21.Based on the above, we can say that we have a scale with good indicators for measuring addiction to social networks in the Chilean population.

Table 1 .
Descriptive analysis of the items

Table 2 .
Correlation between items

Table 6 .
Concurrent validity.DISCUSSIONOur study aimed to analyze and adapt the SMAS, which measures social network addiction.It was found to have good psychometric properties for the Chilean population.Therefore, it can become a suitable tool to measure social network addiction.