Content area
Earlier research focused on investigating the acceptance of educational technologies applied in formal learning settings. Understanding the factors that can lead to the adoption of other dominant technologies in social communication for informal learning is an area that remains under-studied. Moreover, previous literature focused on the use of structural equation modeling (SEM) to predict technology acceptance, whereas the application of data mining algorithms is rare in this direction of research. This study, therefore, aims to (1) propose an integrated framework based on the DeLone and McLean information system model, the diffusion theory, the interactivity theory, the intrinsic motivation theory, and the security perceptions, (2) predict the adoption of TikTok as a learning means in an informal educational space, and (3) compare the performance of data mining techniques and SEM in predicting users’ behavioral intention towards TikTok acceptance. A cross-sectional survey research design is adopted to achieve the research goals. Data from 143 participants are collected and analyzed based on the convenience sampling technique. The partial least square, Support Vector Classifier, and Random Forest techniques are used to identify the predictability of the proposed framework. The findings suggest that the most influential constructs on TikTok adoption are perceived enjoyment, interactivity, security perceptions, and perceived satisfaction. Such factors explain about 83.2% of the variance of behavioral intention towards the adoption of TikTok for informal learning. The study also shows a clear similarity between the findings of SEM and data mining techniques in their overall prediction rate. The key implications of this research are twofold. First, it proposes a modified framework that explains a high variance of TikTok acceptance. Second, in informal learning contexts, particular constructs such as enjoyment, security, interactivity, and satisfaction can affect technology adoption more than information or system quality.
Full text
1. Introduction
The emergence of social media networks (SMNs) has significantly changed how people interact with each other. Such platforms can create new opportunities in different life sectors, including business, economic, healthcare, and education. Governmental and non-governmental organizations adopt SMNs to improve the relationship with their customers and citizens as well as strengthen their engagement (Khlaif & Salha, 2021). Accordingly, several different SMSs have been established such as Facebook, Twitter, YouTube, and TikTok. Each platform has its particular features and focus, i.e., the content of TikTok is intended for short videos. TikTok has been ranked number five in comparison with other SMNs based on the number of active users (Dixon, 2024). Although TikTok was originally established for entertainment, its potential use as an educational tool has also been considered (Xu et al., 2019).
SMNs have a significant effect on pedagogical practices, creating innovative theoretical approaches (Al-Azawei, 2019). Creating innovative theoretical approaches in contemporary education may include suggesting original models or new perspectives in responding to learners’ needs. This may often compete with existing approaches and introduce new ideas. For example, SMNs play a key role in fostering innovative theoretical approaches by introducing new platforms for idea exchange, effective collaboration, and continuous communication (Greenhow & Lewin, 2016). Hence, the popularity of short videos with the aim of achieving a particular learning objective has been raised. In this regard, TikTok represents a potential learning platform in contemporary education because it enables the delivery of short educational videos in a small time period (Khlaif & Salha, 2021). As stated by Fiallos et al. (2021), about 41% of its active users are young people. This should encourage educational institutions to consider such a platform with the aim of enhancing learner engagement. According to the findings of a previous research study (Al-Azawei, 2019), learners tend towards SMNs such as Facebook in education more than formal e-learning platforms. However, this may be more evident for TikTok as it has a perfect format for creating short educational videos and this, in turn, led LearnOnTikTok videos to have more than 72 billion views and hundreds of videos to be uploaded every day (Fiallos et al., 2021). In this context, considering what can drive successful adoption of TikTok as an informal learning setting is necessary. Moreover, TikTok may or may not be implemented successfully in learning settings, particularly if educators aim to integrate this technology with formal learning.
Technology adoption refers to the apparent willingness of users to use the technology for the purposes that it is intended for. Several different technology adoption and success models have been proposed such as the theory of reason action (Ajzen & Fishbein, 1980), the technology acceptance model (Davis, 1986), the unified theory of acceptance and use of technology (Venkatesh et al., 2003), and the DeLone and McLean Information Systems Success model (DeLone & McLean, 2003). This research is grounded on the DeLone and McLean Information Systems Success (D&M-ISS) model due to several reasons. First, it is widely applied in technology success research (Al-Azawei & Al-Azawi, 2021). Second, it is an adaptable model to new contexts and technologies (Petter et al., 2008). Finally, this model can be successfully aligned with emerging learning technologies (Marjanovic et al., 2024). However, this model focuses on quality factors, whereas other constructs that may relate to the current behavior of users are neglected. It should be clear that there is a lack of research on predicting TikTok’s success in the education sector. Moreover, SNSs provide enjoyable and interactive features, so such factors are integrated with the proposed model. However, as reported in earlier research (D’Arcy & Herath, 2011), users may not tend to use an information system or application if they feel that their individual information and privacy are not preserved. This leads to integrating security perceptions to evaluate the success of information systems. Yet, there is limited research on their role in predicting technology adoption (Al-Azawei et al., 2023). This research, therefore, aims to predict the acceptance of TikTok as an informal learning means.
The contributions of this research are threefold. First, it provides a more comprehensive understanding of the factors that can contribute to TikTok’s success in informal education. Second, an integrated model is proposed based on several different theories to show the outcomes of the adoption of SMSs for learning. The integrated theories complement each other. For example, although the D&M-ISS model focuses on quality concepts, it pays little attention to the social effect of technology confusion, its interactive features, intrinsic motivation, and security considerations. Finally, the research compares the performance of structural equation modeling with data mining techniques. To the best of the authors’ knowledge, this research is the first of its kind that integrates five widely adopted theories in technology success research to predict TikTok adoption in informal learning.
The rest of this research is structured as follows. Section 2 presents general background on TikTok technology, whereas Section 3 shows the use of TikTok in education. Section 4 introduces the proposed framework. In Section 5, the research methodology is reported. Section 6 presents the research findings, whereas Section 7 discusses the key research outcomes. Finally, Section 8 highlights the research conclusions.
2. The TikTok Application
TikTok was first launched in China under the name of Douyin in 2016 and internationally under the name of TikTok in 2017. This technology allows 10 min micro videos. Micro videos are the most used type of multimedia on mobile devices. They typically include three features: (1) a brief duration, which makes them simple to share on social media, (2) the subjective expression of micro-video semantics can be facilitated by social attributes such as descriptions, follower counts, and click through rates, and (3) they are created in real time, so videos can show the feelings and emotions of content producers (Guo et al., 2021). TikTok is preferred by its users because the content is concise, expressive, and informative. As TikTok is an application that displays talent content and knowledge, its users enjoy using this application and it may lead to addiction (Y. Yang & Zilberg, 2020). TikTok has also been successfully used in recommendation systems based on artificial intelligence (Grandinetti, 2023). Algorithms of artificial intelligence can accurately learn users’ preferences and provide recommendations based on their past search. Figure 1 shows the most popular SMNs.
Iraq is ranked number one in the Middle East region regarding the number of active TikTok users (IraqiNews, 2023). This is followed by the United Arab Emirates, Saudi Arabia, Kuwait, and Qatar. In Iraq, TikTok has about 23.88 million active users within the following age ranges: 18–24 (65.9%), 25–34 (24.3%), and 35–44 (5.6%) (Start.io, 2024; Kemp, 2023). For all active users in early 2023, 33.7% were female, while 66.3% were male (Kemp, 2023). This may be attributed to the nature of Iraqi culture and religion. In respect to TikTok penetration, Iraq is number three with 100.4%, whereas Saudi Arabia is ranked number one (110.7%), followed by the United Arab Emirates (104.7%). Such statistics should invite educational organizations in Iraq to exploit TikTok features in the learning and teaching process.
3. The Use of TikTok in Learning
According to Mei and Aziz (2022), the key functions of SMSs are interaction, entertainment, knowledge, and education. Hence, such platforms can undoubtedly improve students’ learning experience, but there are still concerns regarding their negative effects (Greenhow & Lewin, 2016). Although TikTok is well-known as an application where users can upload their own videos, participate in trends and display other talents, its use for sharing educational content is beginning to grow (Fiallos et al., 2021). In India, for example, TikTok hosted the EduTok program to revolutionize successful e-learning application (Rahimullah et al., 2022). This was followed by the development of the LearnOnTikTok application by TikTok (Grandinetti, 2023). To create educational content, TikTok works with 800 educational institutions, educators and nonprofit institutions as part of this program. This includes different scientific subjects such as technology and engineering, medicine and healthcare, food and drink, and science/chemistry (Rahimullah et al., 2022).
In line with the previous discussion, TikTok has been widely integrated in education (S. Wang et al., 2024), and this should invite further research to investigate what can affect its successful adoption and implementation. Providing an integrated learning environment by face-to-face and TikTok application may offer a means for learners’ engagement and this, in turn, can lead to improving learning outcomes. However, this cannot be achieved without investigating factors that may affect TikTok successful adoption.
The above-discussed benefits of TikTok use do not mean that this social media application is without limitations in educational contexts. SMSs such as TikTok can also cause (1) learner distraction (Junco, 2012), (2) misinformation (Pennycook & Rand, 2019), and privacy and security issues (Maheshwari & Mac, 2023). Accordingly, SNSs should be used in learning with more consideration on such limitations.
4. The Proposed Framework and Literature Review
In this study, a conceptual framework is proposed, which is grounded on the D&M-ISS model (DeLone & McLean, 2003). However, other constructs that were not considered in the original D&M-ISS model and may have effect on TikTok adoption are integrated. The proposed model includes learner enjoyment, security perceptions, perceived popularity, interactivity, learner satisfaction, information quality, and system quality. Figure 2 depicts the proposed model.
According to DeLone and McLean (1992), several different models on information system success were reviewed. This leads to the proposal of a six-dimensional model. It suggests a relationship among six constructs. These are perceived satisfaction, user impact, information quality, system quality, information system use, and organizational impact. According to Y. S. Wang and Liao (2007), the framework adds two important contributions to earlier literature on information system success. First, it establishes a scheme to categorize the multitude of measures of information system success. It, furthermore, suggests a framework of causal interdependence between factors. Accordingly, this model received researchers’ and scholars’ attention because of its validation and evaluation in different contexts (Petter & McLean, 2009). In 2003, DeLone and McLean reviewed the model and suggested a new construct, namely, service quality (DeLone & McLean, 2003). However, issues were faced by earlier research concerning the validity of this construct (Ramirez-Correa et al., 2017).
Previous literature adopted this model to investigate the success of learning technologies. For example, (Sutiyono et al., 2024) assessed the acceptance of e-learning based on the D&M-ISS model. Similarly, to examine learning management systems (LMS) success, Rulinawaty et al. (2024) also uses this model, confirming its validity in evaluating LMS success. In this context, Widyaningrum et al. (2024) supported its validity to predict LMS success. On the other hand, Alotaibi and Alshahrani (2022) extended the model with two constructs namely, learner and instructor quality in the e-learning context, confirming its validity. In another research study conducted by Rokhman et al. (2022), teacher capability, student capability, and social impact were integrated with the model to evaluate e-learning success. Unlike earlier literature, this present research extends the model based on widely accepted theories in technology adoption. It also examines the effectiveness of the proposed model in informal learning context. This investigation confirms the acceptance of SMNs for learning purposes, considering their content quality, users’ privacy, popularity, intrinsic motivation, and interactive features.
4.1. Enjoyment
Persistence in the workplace, health behavior, and academic performance can be predicted by intrinsic motivation (Fishbach & Woolley, 2022). Hence, activity engagement for its pleasure and inherent satisfaction represents intrinsic motivation. Ryan and Deci (2000) emphasize that enjoyment can fulfill psychological needs such as competence and autonomy as well as enhancing intrinsic motivation by expressing a critical emotional experience. Enjoyment refers to the feelings of pleasure when users adopt technology regardless of the performance resulting from technology use (Dickinger et al., 2008). According to the motivation theory, enjoyment is an influential construct in formalizing people’s acceptance of technology (S. Yang, 2013). TikTok is intended for entertainment more than education, so it is expected that users’ decision to accept it for learning is based on their level of pleasure during the use of this technology. If users enjoy their learning experience on TikTok, they are more likely to continue using the platform for educational purposes. For example, if a user finds a TikTok video on a topic that they are interested in, they will more likely engage with the video by liking, commenting, and sharing it, as well as seeking out similar videos in the future. It is shown that young people prefer mobile communication more than other ways because they feel it is more enjoyable (Dickinger et al., 2008). Enjoyment was found to be a predictor of behavioral intention to adopt mobile learning (Al-Azawei & Alowayr, 2020). It was found that enjoyment affects users’ decision to accept TikTok for learning (Rahimullah et al., 2022). In this research, therefore, the following hypothesis is proposed:
Enjoyment affects users’ decision to accept TikTok for informal learning.
4.2. Security Perceptions
Perceived security refers to perceptions of users regarding the possible security threats of technology use (Belanche-Gracia et al., 2015). According to Maqableh et al. (Maqableh et al., 2021), it is crucial to consider security constructs in investigating technology success. Moreover, Al-Azawei et al. (2023) found that security perceptions were predictors of online learning actual use. This study, therefore, integrates this construct with the proposed model to understand TikTok adoption in informal learning. The rationale behind this integration is that when users perceive that their personal information and data are secure, they are more likely to trust the platform and use it for educational purposes. For example, if technology has strong security measures such as two-factor authentication and encryption, users may feel more comfortable sharing their personal information with the platform, and thus are more likely to use it for learning (Kumar et al., 2018). Moreover, the sensitive nature of educational data may necessitate a high level of security. In earlier literature, perceived security was found to affect users’ decision to adopt technology (Al-Azawei et al., 2023; Kumar et al., 2018). Accordingly, the following hypothesis is proposed:
Security perceptions affect users’ decision to accept TikTok for informal learning.
4.3. Popularity
In the diffusion theory, it was proposed that social systems have a significant effect on people’s decision to accept technology (Rogers, 1995). Accordingly, Rogers (1995) defines social norm as an influential factor on technology adoption. Other scholars have used several different names to refer to the effect of social systems such as “social pressure”, “cultural fashion”, “social atmosphere”, and “bandwagon effects” (Zhu & He, 2002). Looking at statistics as presented in IraqiNews (2023), it can be clear that TikTok becomes very popular in the Iraqi society. Hence, its popularity may affect its adoption in informal learning settings.
Popularity means that when users perceive TikTok as a popular platform for learning, they may be more likely to use it for educational purposes. In earlier literature, it was argued that popularity was the most significant determinant of SMN use (Sun et al., 2014). As such, if a user finds that many people share or use educational content on TikTok, they will be more likely to explore the platform for the same purpose. In short, both the reality of TikTok popularity in Iraq and existing theories may confirm the need for an empirical investigation of the effect of perceived popularity on the acceptance of TikTok. Hence, the following hypothesis is suggested:
Perceived popularity affects users’ decision to accept TikTok for informal learning.
4.4. Interactivity
Interactivity refers to the notion of attracting users to a particular technology. SMNs involve online interaction among users as a key feature of each social network. This interactive communication could be a one-to-one or one-to-many based on different network features (Shipps & Phillips, 2013). In the interactivity theory, it was posited that the quality of communication can affect perceived interactivity (Rafaeli, 1988). Interactive features that are included in each technology should motivate users towards that technology and this, in turn, may influence their decision to adopt it. Earlier studies argued that including interactive features with technology could significantly affect users’ willingness to adopt it (Gan & Balakrishnan, 2016; Shipps & Phillips, 2013). Shipps and Phillips (2013) found that perceived interactivity was a determinant of user satisfaction. Hence, when users are able to interact with the content of TikTok such as asking questions or leaving comments on a video, they may be more likely to use the platform for educational purposes. Hence, the following is assumed in this research:
Interactivity affects users’ decision to accept TikTok for informal learning.
4.5. Perceived Satisfaction
Learner satisfaction is defined as the acquisition of all benefits learners aim to obtain from the learning process, as per their attitudes, behavior, and beliefs (Wu et al., 2010). According to Bolliger and Martindale (2004), learner satisfaction is essential in the adoption of educational technology. Thus, perceived satisfaction is a key construct stemming from performing a learning task, where learning outcomes are derived enjoyably (Al-Azawei & Lundqvist, 2015). Educational institutions should pay special consideration to learner satisfaction. From a commercial perspective, students are similar to customers. Thus, their learning needs should be met. From a learning point of view, students cannot learn properly if they feel that there are environmental or personal barriers preventing the achievement of their objectives. Previous research confirms that learner satisfaction is a predictor of behavioral intention (Al-Azawei et al., 2023; Hussein et al., 2022). Hence, this research suggests the following:
Satisfaction affects users’ decision to accept TikTok for informal learning.
4.6. Information and System Quality
DeLone and McLean (2003) suggested that information and system quality are determinants of behavioral intention. In this research, the former refers to the sufficiency and accuracy of information learners obtained from TikTok, whereas the latter means technical features of TikTok, such as interface design, reliability, stability, and efficiency. These two constructs were also confirmed as strong predictors of intention to use technology in earlier literature (Al-Azawei et al., 2023; Al-Adwan et al., 2022). Accordingly, we suppose the following:
Information quality affects users’ decision to accept TikTok for informal learning.
System quality affects users’ decision to accept TikTok for informal learning.
5. Research Methods
This research focuses on TikTok adoption as a means of informal learning technology based on the perspectives of users in Iraq. Cater-Steel (2004) recommends choosing an appropriate research design in information system research. Hence, a cross-sectional, survey, and quantitative research approach is adopted here because data were collected only once. This study aims to provide a clear understanding of the cause and effect relationship among different constructs, and the collected data are based on numerical values that can be used to investigate the predictability of the proposed model (Saunders et al., 2012). Leedy and Ormrod (2001, p. 183) define the survey research design as “acquiring information about one or more groups of people—perhaps about their characteristics, opinions, attitudes, or previous experiences by asking them questions and tabulating their answers”. According to Saunders et al. (2012), surveys are significant in data collection based on the notion that researchers can focus on estimating constructs with high precision. For collecting research data, surveys are argued as an effective method according to the promise that surveys can help researchers consider estimating constructs with high precision (Saunders et al., 2012). Furthermore, designing a questionnaire based on previous literature can help compare findings (Brislin, 1986).
5.1. The Research Questionnaire and Data Collection
The research instrument includes clear guidance for participants to answer questions properly. It first introduced the aim of this research, clearly demonstrated that the collected data would be used for academic research only; mentioned that participants can withdraw their participation at any time, and that participation is voluntarily. The two parts of the instrument consist of general questions about respondents’ demographic information such as sex, age group, and TikTok use, whereas the other part includes thirty-six questions that are intended to measure the proposed framework constructs. However, two questions from security perceptions and perceived popularity were deleted due to their low outer loading. A Likert scale ranging from 1 for strongly disagree to 5 for strongly agree was used. All questions were adapted from previous research as shown in Appendix A in which each construct in the proposed model was measured based on several items. However, a few changes were made based on the nature of this study. Figure 3 shows the number of items used for each construct.
A Google form was used to collect the research data. It was distributed online via social media, using a nonprobability convenience sampling method. According to Golzar et al. (2022, p. 73), “convenience sampling describes the data collection process from a research population that is effortlessly reachable to the researcher”. A convenience sample is adopted because it simply encompasses individuals who are accessible to a researcher (Pallant, 2013). Although this method may lead to bias, it is cheap and easy to implement, considering the limitations of time and resources. As mentioned in the literature, the key advantages of this technique are (1) less effort is required by researchers in the selection of respondents, (2) low cost in terms of budget and time based on convenience sampling as population is easily accessible, and (3) preparing a list of population characteristics is unrequired because only a specific part is selected (Golzar et al., 2022). Hence, the questionnaire link was distributed to many friends via social media such as Viber and Whatsup and they were asked to share it with other friends. The homogeneity or heterogeneity of the participants cannot be exactly determined because the questionnaire was anonymized. Moreover, no specific information on participants’ work, level of study, and/or organizations or institutions that they belonged to was collected. Participants were asked to rate each factor based on a scale of 1 to 5, with 1 being “strongly disagree” and 5 being “strongly agree”. All questions were determined as mandatory to avoid receiving incomplete responses.
5.2. Data Analysis
The collected data were analyzed using several different methods. Descriptive statistics analysis included mean, standard deviation, and frequency. Pearson correlation was also measured to ensure that the multicollinearity issue was avoided. To confirm the questionnaire properties, item loading, Cronbach’s alpha, composite reliability, average variance extracted, and discernment validity were computed. To measure the cause and effect relationship among the proposed model, partial least square and two data mining techniques, namely, Support Vector Classifier and Random Forest, were implemented. Such analyses were conducted using SPSS version 16.0 and SmartPLS software version 3 as well as Python programming language version 3.11.
SEM is used in this research as it represents the most dominant approach to investigating the cause and effect association among model’s constructs (Al-Azawei et al., 2017). Moreover, SEM is an appropriate approach for the purpose of prediction and development of theories (Chin, 1998). Finally, according to Tobias (1995), this technique is appropriate in determining responses’ behavior (constructs) from the identified variables (independent factors). On the other hand, Support Vector Classifier and Random Forest are widely adopted data mining techniques due to their power in the prediction and classification contexts (Alamri et al., 2020). Support Vector Classifier maximizes the distance that separates the elements of two labels that the elements belong to. When a set of data is used to compute the boundary limit among labels, this is known as the training set, whereas the set of data that is used to test the technique’s performance is called the validation set. Random Forest is based on the Decision Tree technique as it represents a group of several tree predictors. Specifically, each tree relies independently on values of a vector with the same distribution over all trees in a particular forest (AhmedK et al., 2013).
6. Results
The research investigation includes four key steps to understand the relationship among the constructs of the proposed framework. First, descriptive statistics are reported. This is followed by identifying the psychometric properties of the questionnaire. Third, the partial least square technique is applied to highlight the cause and effect association between the independent and dependent features. Finally, two data mining techniques are applied to show the proposed model predictability and highlight the features significance. Overall, 143 respondents voluntarily participated in this research. Table 1 depicts their demographic information. It is clear that most respondents used TikTok more than four hours a day.
6.1. Descriptive Statistics
Table 2 presents the mean, standard deviation, and number of participants who answered questions of each construct. The mean scores of most constructs are higher than the midpoint of 2.5 and ranged from 2.54 to 2.85, except for three factors (enjoyment, satisfaction, and service quality), where their means were close to the midpoint. Standard deviations, on the other hand, ranged from 0.853 to 1.005. This clearly indicates that the spread of values is around the mean.
Table 3 shows the correlation between the constructs based on Pearson tests. There is a significant correlation between the models’ factors, but there is no correlation more than 0.8. This indicates that the features are separated from each other.
6.2. Instrument Properties
Convergent validity refers to the degree to which all items of a particular measure have approximate common variance, whereas discriminant validity means that a factor is distinct from all other factors in a model (Hair et al., 2006). Cronbach’s alpha, Composite Reliability (CR), and Average Variance Extracted (AVE) are measures used to confirm the convergent and discriminant validity of an instrument. The AVE test estimates its approximate association based on the measurement error. The recommended thresholds for Cronbach’s alpha, CR, and AVE are 0.7, 0.5, and 0.7, respectively (Pallant, 2013; Hair et al., 2006). As presented in Table 4 and Table 5, all recommended thresholds were met. After establishing the instrument properties and confirming its internal consistency and constructs’ validity, the PLS model was applied to investigate the proposed hypothesizes. Table 6 reports the outer loading of all items, which confirms that there was no outer loading less than 0.7 to provide further support to the research questionnaire.
6.3. The Results of the Proposed Hypotheses Based on Partial Least Squares
Table 7 and Figure 4 show the path coefficient among all factors of the proposed model. Generally, four hypotheses were confirmed (H1, H2, H5, and H5), whereas H3, H6, and H7 were rejected. H1 supports the relationship between enjoyment and user intention (βenjoyment → user intention = 0.357, p-value < 0.001). Moreover, H2 suggests that the security perception construct was a predictor of user intention (βSecurity Perceptions → user intention = 0.201, p-value < 0.001). As suggested in H4, interactivity also had a strong positive and significant association with behavioral intention (βInteractivity → user intention = 0.212, p-value < 0.001). Finally, perceived satisfaction was a predictor of user intention (βPerceived Satisfaction → user intention = 0.174, p-value < 0.01). On the other hand, perceived popularity was not a determinant of user intention as proposed in H3 (βPerceived Popularity → user intention = 0.050, p-value > 0.05). Similarly, hypotheses H6 and H7 were rejected to suggest that information quality (βInteractivity → user intention = 0.114, p-value > 0.05) and system quality (βInteractivity → user intention = −0.067, p-value > 0.05) were not predictors of user intention.
6.4. The Results of Data Mining Techniques
To confirm the findings of the structural equation modeling and highlight the differences between this method and data mining algorithms, Support Vector and Random Forest techniques were applied. As shown in Table 8, Support Vector Classifier outperforms the other technique where its overall accuracy was 0.72%. However, partial least square suggests higher predictability among the constructs (R2 = 0.83). This means that the independent factors can explain the very good variance of the dependent construct. For further clarity, confusion matrixes are drawn to show the number of the predicted true positive, true negative, false positive, and false negative of each technique (see Figure 5).
To further enhance the predictability of data mining techniques, feature weights were computed, which reflects the possible participation of each feature in predicting user intention. Features’ importance displays how each feature contributed to the prediction task. Figure 6 depicts that enjoyment, security perceptions, information quality, and perceived satisfaction had the highest weights in comparison with other features. Accordingly, only these features were entered into machine learning algorithms to predict the class. The overall performance of both classifiers was improved after computing feature weights where features with low weight were excluded. For the Random Forest technique, its baseline accuracy was 0.65, whereas it was improved to 0.70 after entering only features with the highest weights. Similarly, the baseline performance of Support Vector Classifier was 0.72, but it was enhanced to 0.77 as shown in Table 8.
7. Discussion
The overall outcomes of this research based on the structural equation modeling suggested that users’ enjoyment, security perceptions, interactivity, and perceived satisfaction were predictors of user intention, supporting hypotheses H1, H2, H4, and H5. However, the association among perceived popularity, information quality, and system quality with behavioral intention was not confirmed, rejecting hypotheses H3, H6, and H7.
Pertaining to H1 that suggested a relationship between enjoyment and user intention, the results showed that this construct had the highest influence on TikTok adoption. In fact, this outcome is consistent with earlier research (Al-Azawei & Alowayr, 2020; Arain et al., 2019), confirming the importance of enjoyment in technology use. As such, considering strategies that could motivate learners to use different learning technologies is crucial to ensure that they use such technologies with fun and enjoyment. Security concerns were also a predictor of behavioral intention (H2), supporting the findings of previous literature (Al-Azawei et al., 2023; Farooq et al., 2020). This means that even though learners voluntarily use informal learning technologies, they seek platforms that hide their personal information from unauthorized parties. Moreover, guaranteeing that learning content or personal information will not be modified without a user is also a concern of learners. Such technologies should also be free of errors or technical issues that prevent their effective use. This result confirms the conclusion of Maqableh et al. (2021) that security concerns should be considered in the investigation of technology acceptance.
Another significant predictor of user intentional behavior was interactivity as proposed in H4. It is obvious that SMNs include several different features that match the principles of the interactivity theory such as one-to-one or one-to-many communication. It was found in this research that such interactive features had an influence on users’ willingness to adopt TikTok, confirming the findings of previous research (Gan & Balakrishnan, 2016; Shipps & Phillips, 2013). Thus, high interactive features of TikTok such as leaving comments on particular learning content or asking a question on a video showed a high influence on its adoption as an informal learning technology. As confirmed in other studies (Al-Azawei et al., 2023), satisfaction was a determinant of behavioral intention. Users’ satisfaction covers their beliefs that particular technology could respond to their individual preferences (Dong et al., 2014). DeLone and McLean (2003) also assumed an association between perceived satisfaction and behavioral intention as found in this present research. Hence, to ensure that users will use a particular informal learning technology, their individual needs and preferences should be met. This can be achieved by, for example, personalizing learning content or videos on an SMN platform based on their individual search.
Three hypotheses were rejected in this research. The assumption that there was a relationship between perceived popularity and behavioral intention (H3) was not supported. Although the diffusion theory suggested that social systems can influence users’ decision to adopt technology (Rogers, 1995), this was not the case in all contexts. Previous research also found that social norm was not a predictor of mobile-learning adoption in Iraq (Al-Azawei & Alowayr, 2020). This may indicate that Iraqi people do not build their decision to accept technology according to other preferences. They may use particular technology based on its features that meet their needs and this, in turn, leads to its popularity. Moreover, both information and system quality were not determinants of user intention to adopt TikTok. This means that other features of TikTok had a high influence on its adoption such as enjoyment, security concerns, and interactivity as confirmed in this research. In agreement with this research findings, Jahan et al. (2024) also adopted the D&M-ISS model to examine the adoption of SMNs. The significant influence of information and service quality on social media acceptance was not confirmed. Thus, although D&M ISS had attracted great attention in information system success research, the effect of its original constructs may decrease due to the influence of other factors on technology adoption. Rahimullah et al. (2022) suggested that information and system qualities were determinants of learners’ satisfaction when using TikTok in higher education. This may also support the findings of this present study that both factors had a low effect on behavioral intention.
The research adds key theoretical and practical implications. The former includes the proposal of a modified framework that explains the high variance of TikTok adoption. This study also suggests that in informal learning contexts, enjoyment, security, interactivity, and satisfaction constructs had more effect on technology use than information or system quality. This is because learners use such technologies voluntarily, so they want to have fun and pleasure while using them. The overall findings of structural equation modeling and machine learning techniques are consistent. This supports the practical implications of this research. The former explains about 83% of the variance of behavioral intention, whereas the accuracy outcome of the latter was between 70% and 77%. Support Vector Classifier outperforms Random Forest in both cases before and after weighting the features. This lower performance of machine learning techniques may be attributed to the small dataset used in this research. Another possible reason is the use of a low number of features in the prediction process. However, based on such outcomes, machine learning techniques can also be applied effectively with such a small dataset with no concern of overfitting as their explained accuracy was in line with the findings of structural equation modeling. According to the research findings, the following recommendations are made:
Integrating SMNs as an informal leaning means is becoming important in contemporary education. This is because young people have a greater tendency to use such technologies due to their interactive and enjoyable elements.
SMNs can also be included alongside traditional learning approaches to enhance learning experience.
Teachers and/or lecturers in formal learning settings should be eager to integrate informal learning methods as this may lead to improving learners’ satisfaction and meeting their own needs.
People may tend to use SMNs as an informal learning means based on their included features rather than considering their popularity and widespread use.
8. Conclusions
Based on the assumptions of different theories on technology diffusion and adoption, this research proposed a framework to identify constructs that may affect TikTok adoption as an informal learning platform among Iraqi users. The proposed model integrated the D&M-ISS model with the diffusion theory, the interactivity theory, the intrinsic motivation theory, and the security perceptions. Two prediction techniques were adopted to examine the cause and effect association among the proposed model constructs. These were partial least square and data mining techniques in which their findings suggested that the proposed model can explain from 70% to 83% of TikTok adoption. Based on the overall findings, several conclusions can be drawn, as outlined below:
TikTok represents a supportive platform in teaching and learning.
The interactive features included with this technology play a significant role in its widespread use and adoption.
The high influence of both satisfaction and enjoyment on TikTok adoption could mean that any technology will not be adopted without responding to users’ own needs and providing several different features that make its use fun and enjoyable.
The successful use of a platform highly depends on ensuring that its services are secure.
Although data mining techniques are more appropriate with large datasets, these research findings suggest that such techniques can also be used successfully with small datasets.
Regardless of the significant findings of this research, it is not without its limitations that should be addressed in future research. First, the research instrument was distributed online via a Google form, which might affect the overall understanding of some questions that may need explanation. However, this issue was avoided as much as possible by ensuring clarity of questions. Second, the sample size cannot reflect the whole sample of Iraqi users. Hence, recruiting more participants and from different disciplines is necessary. This can also help in avoiding the overfitting issue in data mining techniques. Third, learners’ perceptions were measured once only, whereas collecting their opinions in two periods can better reflect their perceptions. Finally, the average performance of data mining techniques could be attributed to the low number of features used to predict the class label. Therefore, integrating other constructs with the proposed framework may help explain the unpredicted ratio.
Methodology, A.A.-A. and A.A.; Software, A.A.-A.; Investigation, A.A.-A.; Resources, A.A.-A. and A.A.; Data curation, A.A.-A.; Writing—original draft, A.A.-A.; Writing—review & editing, A.A.; Visualization, A.A.-A.; Supervision, A.A.-A.; Project administration, A.A.; Funding acquisition, A.A. All authors have read and agreed to the published version of the manuscript.
Not applicable.
Informed consent was obtained from all subjects involved in the study.
The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.
The authors thank the University of Babylon, Iraq, and Albaha University, Saudi Arabia, for supporting this research. They also would like to thank all participants for their time and insightful response.
The authors declare no conflicts of interest.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Figure 1. The most popular social networks worldwide as of January 2024, ranked by the number of monthly active users (in millions) (Dixon, 2024).
Demographic information of participants.
| Item | Number |
|---|---|
| Sex | |
| 40 |
| 103 |
| Age Group | |
| 109 |
| 31 |
| 3 |
| Daily Use of TikTok | |
| 6 |
| 1 |
| 34 |
| 92 |
Descriptive statistics.
| Construct | Mean | Std. Deviation | N |
|---|---|---|---|
| Enjoyment | 2.49 | 1.004 | 143 |
| Security Perceptions | 2.57 | 1.005 | 143 |
| User Intention | 2.74 | 0.943 | 143 |
| Perceived Popularity | 2.85 | 0.889 | 143 |
| Service Quality | 2.32 | 0.783 | 143 |
| Information Quality | 2.55 | 0.878 | 143 |
| Interactivity | 2.54 | 0.853 | 143 |
| Satisfaction | 2.43 | 0.861 | 143 |
Pearson correlation.
| SecPerc | UserInt | PercPop | SQ | IQ | Interac | Satis | |
|---|---|---|---|---|---|---|---|
| Enjoy | 0.620 ** | 0.809 ** | 0.667 ** | 0.546 ** | 0.668 ** | 0.614 ** | 0.705 ** |
| SecPerc | 1 | 0.736 ** | 0.581 ** | 0.731 ** | 0.735 ** | 0.641 ** | 0.667 ** |
| UserInt | 1 | 0.745 ** | 0.656 ** | 0.807 ** | 0.784 ** | 0.797 ** | |
| PercPop | 1 | 0.600 ** | 0.731 ** | 0.735 ** | 0.697 ** | ||
| SQ | 1 | 0.745 ** | 0.680 ** | 0.691 ** | |||
| IQ | 1 | 0.842 ** | 0.805 ** | ||||
| Interac | 1 | 0.771 ** |
** Correlation is significant at the 0.01 level (2-tailed).
Cronbach’s alpha, Average Variance Extracted (AVE), and Composite Reliability (CR).
| Construct | Cronbach’s Alpha | rho_A | CR | AVE |
|---|---|---|---|---|
| Enjoyment | 0.923 | 0.924 | 0.945 | 0.812 |
| Information Quality | 0.934 | 0.936 | 0.948 | 0.753 |
| Interactivity | 0.858 | 0.862 | 0.904 | 0.702 |
| Perceived Popularity | 0.774 | 0.775 | 0.869 | 0.689 |
| Perceived Satisfaction | 0.850 | 0.856 | 0.900 | 0.692 |
| Security Perceptions | 0.799 | 0.830 | 0.907 | 0.831 |
| System Quality | 0.877 | 0.881 | 0.911 | 0.671 |
| User Intention | 0.942 | 0.942 | 0.954 | 0.776 |
Discriminant validity.
| Enj | IQ | Inter | PP | PS | SP | SQ | UI | |
|---|---|---|---|---|---|---|---|---|
| Enj | 0.901 | |||||||
| IQ | 0.670 | 0.868 | ||||||
| Inter | 0.617 | 0.844 | 0.838 | |||||
| PP | 0.641 | 0.712 | 0.712 | 0.830 | ||||
| PS | 0.724 | 0.818 | 0.786 | 0.743 | 0.832 | |||
| SP | 0.620 | 0.738 | 0.642 | 0.574 | 0.669 | 0.911 | ||
| SQ | 0.548 | 0.749 | 0.683 | 0.593 | 0.703 | 0.731 | 0.819 | |
| UI | 0.811 | 0.809 | 0.785 | 0.717 | 0.818 | 0.740 | 0.658 | 0.881 |
Enj: enjoyment, IQ: information quality, Inter: interactivity, PP: perceived popularity, PS: perceived satisfaction, SP: Security perceptions, SQ: system quality and UI: user intention.
Outer loadings.
| Item | Enjoy | IQ | Interac | PercPop | Satis | SecPerc | SQ | UserInt |
|---|---|---|---|---|---|---|---|---|
| Enjoy1 | 0.869 | |||||||
| Enjoy2 | 0.930 | |||||||
| Enjoy3 | 0.912 | |||||||
| Enjoy4 | 0.892 | |||||||
| IQ1 | 0.885 | |||||||
| IQ2 | 0.888 | |||||||
| IQ3 | 0.870 | |||||||
| IQ4 | 0.810 | |||||||
| IQ5 | 0.875 | |||||||
| IQ6 | 0.877 | |||||||
| Interac1 | 0.865 | |||||||
| Interac2 | 0.843 | |||||||
| Interac3 | 0.851 | |||||||
| Interac4 | 0.790 | |||||||
| PercPopu1 | 0.816 | |||||||
| PercPopu2 | 0.844 | |||||||
| PercPopu3 | 0.829 | |||||||
| SQ1 | 0.800 | |||||||
| SQ2 | 0.803 | |||||||
| SQ3 | 0.878 | |||||||
| SQ4 | 0.782 | |||||||
| SQ5 | 0.830 | |||||||
| Satis1 | 0.850 | |||||||
| Satis2 | 0.876 | |||||||
| Satis3 | 0.854 | |||||||
| Satis4 | 0.742 | |||||||
| SecPerc1 | 0.889 | |||||||
| SecPerc2 | 0.933 | |||||||
| UserInt1 | 0.856 | |||||||
| UserInt2 | 0.886 | |||||||
| UserInt3 | 0.903 | |||||||
| UserInt4 | 0.886 | |||||||
| UserInt5 | 0.871 | |||||||
| UserInt6 | 0.884 |
The research findings based on PLS.
| Hypothesis | T Values | p Values | Beta | Results |
|---|---|---|---|---|
| H1: Enjoyment -> User Intention | 5.844 | <0.001 | 0.357 | Accepted |
| H2: Security Perceptions -> User Intention | 3.370 | <0.001 | 0.201 | Accepted |
| H3: Perceived Popularity -> User Intention | 0.769 | 0.442 | 0.050 | Rejected |
| H4: Interactivity -> User Intention | 3.587 | <0.001 | 0.212 | Accepted |
| H5: Perceived Satisfaction -> User Intention | 2.610 | <0.009 | 0.174 | Accepted |
| H6: Information Quality -> User Intention | 1.453 | 0.147 | 0.114 | Rejected |
| H7: System Quality -> User Intention | 1.056 | 0.291 | −0.067 | Rejected |
The evaluation matrix of the data mining techniques before and after computing features’ weights.
| Before Computing Features’ Weights | ||||
| Technique | Precision | Recall | F1-Score | Accuracy |
| Support Vector Classifier | 0.80 | 0.80 | 0.80 | 0.72 |
| Random Forest | 0.61 | 0.92 | 0.73 | 0.65 |
| After Computing Features’ Weights | ||||
| Technique | Precision | Recall | F1-Score | Accuracy |
| Support Vector Classifier | 0.80 | 0.80 | 0.80 | 0.77 |
| Random Forest | 0.80 | 0.80 | 0.80 | 0.70 |
Appendix A. Research Questionnaire
| Items | Construct | References |
| Enjoyment (JOY) | Adapted from | |
| JOY1 | I feel motivated when I use the TikTok app to access higher educational content | |
| JOY2 | I enjoy accessing higher educational content on the TikTok app | |
| JOY3 | I find the time I spend looking at higher educational content on the Tiktok application fun | |
| JOY4 | I find it entertaining to use the TikTok app to access higher educational content | |
| Security Perceptions (SPs) | Adapted from | |
| SP1 | I am concerned about my privacy when using Tiktok | |
| SP2 | I worry that my personal information may be used for purposes other than what I intended on Tiktok | |
| SP3 | I feel that using social media increases my risk of identity theft | |
| Perceived Popularity (PP) | Adapted from | |
| PP1 | How often do you see other people talking about or sharing content related to this product/brand on social media platforms? | |
| PP2 | Based on what you see on social media, how popular do you perceive this product/brand to be? | |
| PP3 | How important is it to you that the products/brands you use are popular on social media? | |
| PP4 | How likely are you to purchase a product/brand that is perceived to be popular on social media? | |
| Interactivity (INT) | Adapted from | |
| INT1 | I feel the TikTok app can speed up my understanding compared to self-study | |
| INT2 | I feel the TikTok app can respond according to my expectations | |
| INT3 | The TikTok app can facilitate communication between higher educational content creators and other TikTok users | |
| INT4 | The TikTok app makes it easy for me to exchange opinions with fellow users | |
| Satisfaction (Sat) | Adapted from | |
| SAT1 | I feel satisfied using the TikTok app to access higher educational content | |
| SAT2 | I feel happy using the TikTok app to access higher educational content. | |
| SAT3 | I am happy with TikTok’s features (search, hashtags, comments, etc.) with regard to obtaining higher educational content. | |
| SAT4 | I am satisfied with my decision to obtain higher educational content using the TikTok app | |
| Information Quality (IQ) | Adapted from | |
| IQ1 | I feel that the higher educational content I obtain from the TikTok app is trustworthy | |
| IQ2 | I feel the higher educational content on the TikTok app is qualified | |
| IQ3 | I feel the higher educational content on the TikTok app covers many areas (finance, health, language, etc.) | |
| IQ4 | I find the higher educational content I obtain from the TikTok app easy to understand. | |
| IQ5 | TikTok provides information that is easy to understand | |
| IQ6 | TikTok provides up-to-date information | |
| System Quality (SQ) | Adapted from | |
| SQ1 | I feel the TikTok application is not slow | |
| SQ2 | I feel the TikTok app can be used whenever I need it | |
| SQ3 | I feel the TikTok app works as expected | |
| SQ4 | I find the TikTok app easy to use | |
| SQ5 | I feel the TikTok app provides the features I need | |
| User Intention (UserInt) | Adapted from | |
| UserInt1 | I intend to use TikTok in learning in the future | |
| UserInt2 | I will always try to use TikTok in my daily learning | |
| UserInt3 | I plan to use TikTok in learning in the future | |
| UserInt4 | I will recommend other users to use TikTok to learn | |
| UserInt5 | Assuming I have access to TikTok, I intend to use it to learn | |
| UserInt6 | Given that I have access to TikTok, I predict that I would use it to learn | |
References
AhmedK, A.; Aljahdali, S.; Naimatullah Hussain, S. Comparative Prediction performance with support vector machine and random forest classification techniques. International Journal of Computer Applications; 2013; 69, pp. 12-16. [DOI: https://dx.doi.org/10.5120/11885-7922]
Ajzen, I.; Fishbein, M. Understanding Attitudes and predicting social behaviour; Prentice-Hall, Inc.: 1980.
Al-Adwan, A. S.; Nofal, M.; Akram, H.; Albelbisi, N. A.; Al-Okaily, M. Towards a sustainable adoption of e-learning systems: The role of self-directed learning. Journal of Information Technology Education: Research; 2022; 21, pp. 245-267. [DOI: https://dx.doi.org/10.28945/4980]
Alamri, L. H.; Almuslim, R. S.; Alotibi, M. S. Predicting student academic performance using support vector machine and random forest. 2020 3rd International Conference on Education Technology Management (ICETM 2020); London, UK, December 17–19; 2020; pp. 100-107. [DOI: https://dx.doi.org/10.1145/3446590.3446607]
Al-Azawei, A. What drives successful social media in education and E-learning? A comparative study on Facebook and moodle. Journal of Information Technology Education: Research; 2019; 18, pp. 253-274. [DOI: https://dx.doi.org/10.28945/4360] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39910090]
Al-Azawei, A.; Abdullah, A. A.; Mohammed, M. K.; Abod, Z. A. Predicting online learning success based on learners’ perceptions: The integration of the information system success model and the security triangle framework. International Review of Research in Open and Distance Learning; 2023; 24,
Al-Azawei, A.; Al-Azawi, R. Evaluating facebook success in Iraq: An extension of the DeLone and McLean’s model of information systems success (ISS). Journal of Physics: Conference Series; 2021; 1804,
Al-Azawei, A.; Alowayr, A. Predicting the intention to use and hedonic motivation for mobile learning: A comparative study in two Middle Eastern countries. Technology in Society; 2020; 62, 101325. [DOI: https://dx.doi.org/10.1016/j.techsoc.2020.101325]
Al-Azawei, A.; Lundqvist, K. Learner differences in perceived satisfaction of an online learning: An extension to the technology acceptance model in an arabic sample. Electronic Journal of E-Learning; 2015; 13,
Al-Azawei, A.; Parslow, P.; Lundqvist, K. Investigating the effect of learning styles in a blended e-learning system: An extension of the technology acceptance model (TAM). Australasian Journal of Educational Technology; 2017; 33,
Alotaibi, R. S.; Alshahrani, S. M. An extended DeLone and McLean’s model to determine the success factors of E-learning platform Raed. PeerJ Computer Science; 2022; 8, e876. [DOI: https://dx.doi.org/10.7717/peerj-cs.876]
Alowayr, A.; Al-Azawei, A. Predicting mobile learning acceptance: An integrated model and empirical study based on the perceptions of higher education students. Australasian Journal of Educational Technology; 2021; 37,
Arain, A. A.; Hussain, Z.; Rizvi, W. H.; Vighio, M. S. Extending UTAUT2 toward acceptance of mobile learning in the context of higher education. Universal Access in the Information Society; 2019; 18,
Belanche-Gracia, D.; Casaló-Ariño, L. V.; Pérez-Rueda, A. Determinants of multi-service smartcard success for smart cities development: A study based on citizens’ privacy and security perceptions. Government Information Quarterly; 2015; 32,
Bolliger, D. U.; Martindale, T. Key factors for determining student satisfaction in online courses. International Journal on E-Learning; 2004; 3,
Brislin, R. W. Lonner, W. J.; Berry, J. W. The wording and translation of research instruments. Field methods in cross-cultural research; Sage Publications, Beverly Hills: 1986; pp. 137-164.
Cater-Steel, A. An evaluation of software development practice and assessment-based process improvement in small software; Griffith University: 2004.
Chin, W. W. Issues and opinion on structural equation modeling. MIS Quarterly; 1998; 22,
D’Arcy, J.; Herath, T. A review and analysis of deterrence theory in the IS security literature: Making sense of the disparate findings. European Journal of Information Systems; 2011; 20,
Davis, F. D. A technology acceptance model for empirically testing new end-user information systems: Theory and results. Doctorol Thesis; Wayne State University: 1986; Available online: http://dspace.mit.edu/handle/1721.1/15192 (accessed on 2 May 2014).
DeLone, W. H.; McLean, E. R. Information systems success: The quest for the dependent variable. Information Systems Research; 1992; 3,
DeLone, W. H.; McLean, E. R. The Delone and Mclean model of information systems success: A ten-year update. Journal of Management Information Systems; 2003; 19,
Dickinger, A.; Arami, M.; Meyer, D. The role of perceived enjoyment and social norm in the adoption of technology with network externalities. European Journal of Information Systems; 2008; 17,
Dixon, S. J. Global social networks ranked by number of users 2024; 2024; Available online: https://www.statista.com/statistics/272014/global-social-networks-ranked-by-number-of-users/ (accessed on 3 December 2024).
Dong, T. P.; Cheng, N. C.; Wu, Y. C. J. A study of the social networking website service in digital content industries: The Facebook case in Taiwan. Computers in Human Behavior; 2014; 30, pp. 708-714. [DOI: https://dx.doi.org/10.1016/j.chb.2013.07.037]
Farooq, A.; Ahmad, F.; Khadam, N.; Lorenz, B.; Isoaho, J. The impact of perceived security on intention to use e-learning among students. IEEE 20th International Conference on Advanced Learning Technologies, ICALT 2020; Tartu, Estonia, July 6–9; 2020; pp. 360-364. [DOI: https://dx.doi.org/10.1109/ICALT49669.2020.00115]
Fiallos, A.; Fiallos, C.; Figueroa, S. Tiktok and education: Discovering knowledge through learning videos. 2021 8th International Conference on eDemocracy and eGovernment, ICEDEG 2021; Bern, Switzerland, June 18–20; 2021; pp. 172-176. [DOI: https://dx.doi.org/10.1109/ICEDEG52154.2021.9530988]
Fishbach, A.; Woolley, K. The structure of intrinsic motivation. Annual Review of Organizational Psychology and Organizational Behavior; 2022; 9, pp. 339-363. [DOI: https://dx.doi.org/10.1146/annurev-orgpsych-012420-091122]
Gan, C. L.; Balakrishnan, V. An empirical study of factors affecting mobile wireless technology adoption for promoting interactive lectures in higher education. International Review of Research in Open and Distance Learning; 2016; 17,
Golzar, J.; Noor, S.; Tajik, O. Convenience Sampling. International Journal of Education & Language Studies Using; 2022; 1,
Grandinetti, J. Examining embedded apparatuses of AI in Facebook and TikTok. AI and Society; 2023; 38,
Greenhow, C.; Lewin, C. Social media and education: Reconceptualizing the boundaries of formal and informal learning. Learning, Media and Technology; 2016; 41,
Guo, J.; Nie, X.; Ma, Y.; Shaheed, K.; Ullah, I.; Yin, Y. Attention based consistent semantic learning for micro-video scene recognition. Information Sciences; 2021; 543, pp. 504-516. [DOI: https://dx.doi.org/10.1016/j.ins.2020.05.064]
Hair, J. F.; Black, W. C.; Babin, B. J.; Anderson, R. E.; Tatham, R. L. Multivariate data analysis; 6th ed. Pearson Prentice Hall: 2006.
Hussein, M. H.; Ow, S. H.; Al-Azawei, A.; Ibrahim, I. What drives s tudents’ successful reuse of online learning in higher education? A case of Google Classroom. Australasian Journal of Educational Technology; 2022; 38,
IraqiNews,. Iraq ranks as one of the world’s most active TikTok and Snapchat users; 2023; Available online: https://www.iraqinews.com/iraq/iraq-ranks-as-one-of-the-worlds-most-active-tiktok-and-snapchat-users/#google_vignette (accessed on 6 February 2024).
Jahan, N.; Azam, S.; Hossain, A. Examining social networking sites users’ benefits using the DeLone and McLean information system success model. Journal of Information Science Theory and Practice; 2024; 12,
Junco, R. Too much face and not enough books: The relationship between multiple indices of facebook use and academic performance. Computers in Human Behavior; 2012; 28,
Kemp, S. DIGITAL 2023: IRAQ; 2023; Available online: https://datareportal.com/reports/digital-2023-iraq (accessed on 6 February 2024).
Khlaif, Z. N.; Salha, S. Using TikTok in education: A form of micro-learning or nano-learning?. Interdisciplinary Journal of Virtual Learning in Medical Sciences; 2021; 12,
Kumar, A.; Adlakaha, A.; Mukherjee, K. The effect of perceived security and grievance redressal on continuance intention to use M-wallets in a developing country. International Journal of Bank Marketing; 2018; 36,
Leedy, P. D.; Ormrod, J. E. Practical research: Planning and design; 7th ed. Merrill Prentice Hall: 2001.
Maheshwari, S.; Mac, R. Driver’s licenses, addresses, photos: Inside how TikTok shares user data; 2023; Available online: https://www.nytimes.com/2023/05/24/technology/inside-how-tiktok-shares-user-data-lark.html (accessed on 9 March 2025).
Maqableh, M.; Hmoud, H. Y.; Jaradat, M. Integrating an information systems success model with perceived privacy, perceived security, and trust: The moderating role of Facebook addiction. Heliyon; 2021; 7, pp. 1-15. [DOI: https://dx.doi.org/10.1016/j.heliyon.2021.e07899]
Marjanovic, U.; Mester, G.; Marjanovic, B. M. Assessing the success of artificial intelligence tools: An evaluation of chatgpt using the information system success model. Interdisciplinary Description of Complex Systems; 2024; 22,
Mei, K. L.; Aziz, A. A. Students’ perception on using Tiktok application as an english learning tool. International Journal of Academic Research in Progressive Education and Development; 2022; 11,
Pallant, J. SPSS survival manual: A step by step guide to data analysis using IBM SPSS; 5th ed. McGraw Hill: 2013.
Pelletier, M. J.; Horky, A. B. Exploring the facebook like: A product and service perspective. Journal of Research in Interactive Marketing; 2015; 9,
Pennycook, G.; Rand, D. G. Fighting misinformation on social media using crowdsourced judgments of news source quality. Proceedings of the National Academy of Sciences; 2019; 116,
Petter, S.; DeLone, W.; McLean, E. Measuring information systems success: Models, dimensions, measures, and interrelationships. European Journal of Information Systems; 2008; 17, pp. 236-263. Available online: https://link.springer.com/content/pdf/10.1057/ejis.2008.15.pdf (accessed on 17 November 2014). [DOI: https://dx.doi.org/10.1057/ejis.2008.15]
Petter, S.; McLean, E. R. A meta-analytic assessment of the DeLone and McLean IS success model: An examination of IS success at the individual level. Information and Management; 2009; 46,
Rafaeli, S. Interactivity From new media to communication. Advancing communication science: Merging mass and interpersonal process; Sage: 1988; pp. 110-134.
Rahimullah, N. A.; Damayanti, S. B.; Izra, A. A.; Handayani, P. W. Assessing the factors influencing users accessing higher education content on TikTok. Cogent Education; 2022; 9,
Ramirez-Correa, P. E.; Javier Rondan-Cataluña, F.; Arenas-Gaitán, J.; Alfaro-Perez, J. L. Moderating effect of learning styles on a learning management system’s success. Telematics and Informatics; 2017; 34, pp. 272-286. [DOI: https://dx.doi.org/10.1016/j.tele.2016.04.006]
Rogers, E. M. Diffusion of Innovations: Modifications of a model for telecommunications. Die diffusion von innovationen in der telekommunikation; Springer: 1995; pp. 25-38.
Rokhman, F.; Mukhibad, H.; Hapsoro, B. B.; Nurkhin, A. E-learning evaluation during the COVID-19 pandemic era based on the updated of Delone and McLean information systems success model.pdf. Cogent Education; 2022; 9,
Rulinawaty,; Samboteng, L.; Purwanto, A. J.; Kuncoro, S.; Jasrial,; Tahilili, M. H.; Efendi, Y.; Karyana, A. Investigating the influence of the updated DeLone and McLean information system (IS) success model on the effectiveness of learning management system (LMS) implementation. Cogent Education; 2024; 11,
Ryan, R. M.; Deci, E. L. Self-Determination Theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychological Associatin; 2000; 55,
Saunders, M.; Lewis, P.; Thornhill, A. Research methods for business students; 6th ed. Pearson Education Limited: 2012.
Shipps, B.; Phillips, B. Social networks, interactivity and satisfaction: Assessing socio-technical behavioral factors as an extension to technology acceptance. Journal of Theoretical and Applied Electronic Commerce Research; 2013; 8,
Start.io. Tiktok users in Iraq; 2024; Available online: https://www.start.io/audience/tiktok-users-in-iraq (accessed on 6 February 2024).
Sun, N.; Rau, P. P. L.; Ma, L. Understanding lurkers in online communities: A literature review. Computers in Human Behavior; 2014; 38, pp. 110-117. [DOI: https://dx.doi.org/10.1016/j.chb.2014.05.022]
Sutiyono, A.; Bakhri, S.; Ningsih, D. R.; Syukur, F. Assessing E-learning systems using the DeLone & McLean and technology acceptance models. Jurnal Penelitian Pendidikan; 2024; 10,
Tobias, R. D. An introduction to partial least squares regression. Twentieth Annual SAS Users Group International Conference; Orlando, FL, USA, April 2–5; 1995; Available online: https://stats.oarc.ucla.edu/wp-content/uploads/2016/02/pls.pdf (accessed on 7 December 2014).
Venkatesh, V.; Morris, M. G.; Davis, G. B.; Davis, F. D. User acceptance of information technology: Toward a unified view. Source: MIS Quarterly; 2003; 27,
Wang, S.; Sun, Z.; Li, M.; Zhang, H.; Metwally, A. H. S. Leveraging TikTok for active learning in management education: An extended technology acceptance model approach. The International Journal of Management Education; 2024; 22,
Wang, Y. S.; Liao, Y. W. Assessing eGovernment systems success: A validation of the DeLone and McLean model of information systems success. Government Information Quarterly; 2007; 25,
Widyaningrum, T.; Sholihah, Q.; Haryono, B. S. The Delone and McLean information system success model: Investigating user satisfaction in learning management system. Journal of Education Technology; 2024; 8,
Wu, J.-H.; Tennyson, R. D.; Hsia, T.-L. A study of student satisfaction in a blended E-learning system environment. Computers & Education; 2010; 55,
Xu, L.; Yan, X.; Zhang, Z. Research on the causes of the “Tik Tok” App becoming popular and the existing problems. Journal of Advanced Management Science; 2019; 7,
Yang, S. Understanding undergraduate students’ adoption of mobile learning model: A perspective of the extended UTAUT2. Journal of Convergence Information Technology; 2013; 8,
Yang, Y.; Zilberg, I. E. Understanding young adults’ TikTok usage. Communication Research; 2020; 2,
Zhu, J. J. H.; He, Z. Perceived characteristics, perceived needs, and perceived popularity: Adoption and use of the internet in China. Communication Research; 2002; 29,
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.