Social Media Engagement and Student’s Intention in Indonesian Higher Education Using Uniﬁed Theory of Acceptance and Use of Technology

Understanding the motivations behind the use of social media in higher education is crucial for assessing its potential beneﬁts and challenges. However, examining the contribution of social media on collaborative learning within the context of Indonesian universities holds signiﬁcance due to the country’s growing digital landscape and increasing adoption of social media platforms. This research aimed to analyze proposed collaborative learning models involving social media use among Indonesian University students. The proposed models are constructed based on the Technology Acceptance Model (TAM) and the Uniﬁed Theory of Acceptance and Use of Technology (UTAUT). The quantitative method used was analyzing primary data from 143 private university students in Indonesia. Data were collected using a 5-point Likert scale self-reported questionnaire of Internet Anxiety, Habit, and Performance Expectancy and Behavior Intention as well as Social Media Engagement to mediate collaborative learning. The result of this study was that Social Media Engagement and Behavior Intention signiﬁcantly inﬂuence Performance Expectancy and Habit. There was also a signiﬁcant inﬂuence of Internet Anxiety on behavioral intention. Thus, Collaborative Learning is signiﬁcantly inﬂuenced by Social Media Engagement. These results provided insight into developing effective strategies for integrating social media into higher education.


INTRODUCTION
Ì 607 themselves creatively, or enhance their sense of belonging within the academic community. Additionally, social media can provide opportunities for personalized learning experiences, allowing students to tailor their educational journey to their individual needs and interests.
Examining the impact of social media on collaborative learning within the context of Indonesian universities holds significance due to the country's growing digital landscape and increasing adoption of social media platforms. The purpose of this study was to analyze the impact of the use of social media on collaborative learning in the environment of universities in Indonesia. Collaborative learning is an instructional approach emphasizing group work, cooperation, and active participation among students to enhance their understanding and learning outcomes [11]. Integrating social media platforms into higher education has provided students new avenues for communication, information sharing, and collaboration [12]. By leveraging social media tools, students can engage in collaborative activities such as group discussions, knowledge sharing, peer feedback, and project coordination, regardless of geographical constraints or time limitations [13]. This study aims to shed light on how social media usage influences collaborative learning outcomes, identifying both the positive and negative effects. The novelty of this research is that this research is focused on the purpose of using social media in education schemes to explore more regarding collaborative learning driven by social media engagement and general use goals. Social media engagement represents the strength of personal relationships, values adopted, and considerations in the use of social media. The social media engagement study from the UGT perspective was carried out by Conffeto [23], which viewed the influence of the categories of content from social media on social media engagement.
Moreover, by extrapolating internal and external aspects as an engagement predictor, this study elaborates on both psychological and technological aspects to predict social media engagement. It refers to the psychological aspects with the proxies are internet-anxiety and habit and technological factors from social media, that is, performance expectancy. Habit refers to students' habit in the use of e-learning as a determinant variable used in this study, which is expected to show a significant effect on the behavioral intention that was confirmed by Ciric [24]. Another variable added to behavior models is anxiety, that according to Wei [25], anxiety, in terms of computer use, will reduce computer self-efficacy and further decrease the ease of use of computers. According to Songkhla [26], the individual characteristics of technology users, including computer anxiety, will affect the ease of use of information technology in the context of web-based learning. According to Yoo [27], Internet self-efficacy focuses on what someone believes he can achieve online now or in the future. However, as a big picture, three determinants are taken from information technology adoption models, particularly the Technology Acceptance Model and the Unified Theory of Acceptance and Use of Technology.

RESEARCH METHOD
This research is quantitative. The primary data were collected using a 5-point Likert scale of Internet Anxiety, Habit, Performance Expectancy, Behavior Intention, Social Media Engagement, and Collaborative Learning. The research data collection method uses purposive sampling. Students using e-learning as well as possessing social media accounts were respondent criteria of this study, i.e., 143 students from two private Universities in Indonesia. The measurement scale used in this research was the 5-Likert scale. At the research development stage, the instrument used was Cronbach alpha for the reliability test and Kaisser-Meyer-Olkin for the validity test.
The research model is based on UTAUT models modified by elaborating on social media use. The analysis of the research model applied a quantitative approach, and the Structural Equation Model proved the hypotheses. Three exogenous variables were Internet anxiety, habit, and Performance expectancy. Four questions measured Internet anxietythree questions for and three for performance expectancy. As a mediating variable, both social media engagement and behavior intention is measured using three items questions. While Social Media Engagement is constructed by the purpose and value of using social media, the Behaviour intention variable refers to the student's intention to plan, try and keep using the system. Four questions measured collaborative learning as an endogenous variable. In general, the research flow will be depicted in Figure 1. The hypothesis testing uses Structural Equation Model (SEM). According to Okaily [28], SEM is a statistical methodology that conducts a multivariate analysis of multicausal relationships between various independent phenomena based on observations. SEM provides a better empirical approach to testing theoretical models by involving measurement and structural models in one analysis [29].

3.
RESULT AND ANALYSIS

Characteristics of Respondents
Most of the respondents were women, as many as 98 people (68.5%) with an average age of 20 years. Most of the respondents (86%) did not subscribe to the internet at home but subscribed to an internet connection via smartphone (81.1%). The percentage of a computer person is 31.5%, with an average experience of using the internet and social media at 5.11 and 4.73 years. The average student allowance is IDR 441776, with an average cellular credit expense of IDR 43750. Three social media that are always or often used are Whatsapp (92.3%), followed by Facebook (59.5%) and Instagram (56.0%).

Measurement Model
The measurement model is conducted as the first stage in Structural Equation Model, using Confirmatory Factor Analysis (CFA) to measure the model, and its results are presented in Table 1. Three items were omitted, namely NA1, AN2, and CL4, because they had a low loading factor which caused a low value of Average Variance Extracted (AVE) and Composite Reliability (CR). AVE shows convergent validity, while CR shows the internal consistency size of the measurement model. The measurement model, in general, is still acceptable with AVE values above 0.5, except for Habit, CR is greater than 0.7 [29].

Structural Model
The second stage of SEM analysis is testing the structural model used to test the research hypothesis. The structural model for the standardized model is presented in Figure 2. Before testing the hypothesis, the structural model presented in Figure 1 above was evaluated for the suitability of the model by using a number of Goodness of Fit (GOF) measures, whose results are presented in Table  2.  The measurement model, in general, is relatively acceptable, seen from the parameter values better than the threshold stated by Okaily [28], namely CMIN / DF smaller than 2, IFI / TLI / CFI greater than 0.9, and RMSEA smaller than 0.08. Thus, the model analysis continues to test hypotheses whose results are presented in Table 3. Performance expectations and habits significantly influence social media engagement, while the influence of internet anxiety is not proven. The higher the students' usefulness and habits, the higher the social media engagement. The influence of habit on social engagement is greater than performance expectancy. Habits, performance expectancy, and internet anxiety significantly affect behavioral intention. The influence of habit also influences behavioral intention more than expectancy and internet anxiety performance. Especially for internet anxiety, the higher the anxiety in internet use will reduce the behavioral intention. At the same time, the influence of performance expectancy and habit shows the direction of a positive relationship. Only social media engagement significantly affects collaborative learning in a positive direction, while the influence of behavioral intention on collaborative learning Social Media Engagement . . . (Matrissya Hermita)

ISSN: 2476-9843
is not proven. The effect of significant performance expectancy on the intention to use social media follows the results of research by Yuan [20] for the context of using social media for business purposes. The difference lies in other determinants, which in this study used internet anxiety and habit, while Nawi et al. used three other exogenous variables: risk, trust, and enjoyment. This study also supports the results of Mensah's [21] study, which used all exogenous variables in UTAUT, and the results were that the same performance expectancy had a significant effect on the intention to use social media. Nilasari [30], which uses two determinants of UTAUT, namely performance expectancy and effort expectancy, also shows that only performance expectancy significantly affects social media use.
The effect of internet anxiety on the use of social media is in accordance with the results of Shongkla's [26] study of computer use cases. Internet anxiety can lead to diverse levels of technology use, including social media. In this study, internet anxiety only significantly affected social engagement in a positive direction. Respondents belonging to the younger generation, in general, do not show concern about using the internet because all of them have had considerable experience using it, with an average of 5 years. Younger students have more experience using social media, but older students, in this case, who have entered the advanced level, are more interested in using media for learning purposes [8]. Understanding the motivations behind the use of social media in higher education is crucial for assessing its potential benefits and challenges. The Uses and Gratifications Theory (UGT), pioneered by Elihu Katz, provides a valuable framework for examining individuals' motivations for media consumption [7]. According to UGT, individuals actively select and use media to fulfill specific social and psychological needs [9]. In the context of social media, users have their own goals and motivations, influenced by factors such as age, gender, and social environment [10].
The use of social media for learning, in this case, collaborative learning, proves that UGT can also be applied to using social media for learning motives. This study is in line with the research of Froget [13], which shows that the use of Facebook for discussion purposes will increase the intensity of the use of social media. The discussion motive is in line with collaborative learning, one of the learning activities discussed between fellow social media users in the context of learning. Tehmina [7] mentions the use of social media from other perspectives on social media engagement, namely, student attachment will influence their identity in universities and student communities, representing a sense of belonging and pride. Social media engagement in this study points to the closeness of relationships and values that support using social media, although it is not fully utilized for learning purposes. This is as stated by Ghosh [31], that not all students consider social media useful for communication and learning. The results of this study indicate that social media has great potential to support learning. Some students informally have used social media to support their lecture activities. They are also classified as generations who have experience in internet use and are accustomed to using social media. Students generally consider themselves capable of using better technology resources than those specifically developed to support academic activities [1]. This statement implies that universities do not need to make a special information system for collaborative learning, for example, as a feature of e-learning systems, but use social media only formally. The findings of this research can contribute to the development of effective strategies for integrating social media into higher education, maximizing its potential for enhancing collaborative learning experiences and improving educational outcomes for students in Indonesia and beyond.

CONCLUSION
Personal characteristics still show a greater influence on social engagement than social media technology factors as measured by performance expectancy, while the anxiety internet does not affect social media engagement. It is related to the respondents belonging to the younger generation who tend to be literate on the internet and social media and are accustomed to using them daily. The intention to continue using social media is influenced by performance expectancy, habits, and internet anxiety, and habits remain the biggest influence. The significant influence of social media engagement on collaborative learning shows that students are more encouraged by the closeness of relationships, values, and considerations in using social media for learning purposes. The intention of continuing to use social media does not guarantee that students can make full use of social media for learning. It shows that the purpose and motivation for using social media are still relatively diverse according to the characteristics and lifestyles of students. The ability to use social media positively needs to be campaigned, especially by universities, to integrate social media with the teaching and learning process. In terms of students, they need to improve their ability to filter information on positive social media. The negative sides of social media need to be alleviated while enhancing the positive results. Universities, especially lecturers, can also create special groups on social media, such as class groups, which can be used to support learning. Participants stated that social media skills should be taught in optional classes covering generic competencies.

Ì 611
In conclusion, using social media in higher education, guided by the Uses and Gratifications Theory, presents a promising avenue for promoting collaborative learning. By understanding the motivations behind students' social media usage and examining its impact within the Indonesian university environment, this study aims to provide valuable insights that can inform educators, policymakers, and researchers to leverage social media effectively for educational purposes.
The main limitation of this study is the relatively small number of samples that might influence the goodness of fit of the structural model. Further research with larger samples needs to be continued to explore further the use of social media for learning purposes, specifically the moderating effects of gender and the experience of using social media. The two moderators included four moderators included in the UTAUT model. The difference between social media platforms, which in this study is dominated by Whatsapp, Facebook, and Instagram, also needs to be further investigated to find out which platforms are most suitable to support learning. From the perspective of behavioral models of technology use, for example, TAM and UTAUT, student perceptions of performance expectancy and effort expectancy from UTAUT may vary, and their impact on the objectives and motivation of media use can vary depending on the technological aspects of the various platforms.

DECLARATIONS
AUTHOR CONTIBUTION The first and second authors contributed to designing the research framework, infield studies, data collection instrument preparation, drafting, and finalizing the manuscript. The third and Fourth authors contributed to data processing and statistical analysis. The fifth author contributed to field studies and data collection.