Motivators for University of Professional Studies Accra Students to Adopt a Learning Management System in Ghana

: The use of a learning management system ( LMS ) is believed to be signi ﬁ cant for students ’ academic performance, but students ’ attitudes regarding its use are divided. Therefore, the purpose of this study is to apply a modi ﬁ ed version of the technology acceptance model to determine the motivators for University of Professional Studies Accra students ’ behavioural intentions ( BIs ) and actual use of LMSs using a convenient sampling technique to elicit data from ﬁ rst - year students in the faculty of information technology and communication studies. The investigation included a total of 188 samples. The study used structured equation modelling partial least squares to analyse the data. Speci ﬁ cally, the study employed the importance - performance map analysis to establish the fac - tors that are important to students ’ actual use of an LMS. The study ’ s ﬁ ndings suggest that REL and output quality signi ﬁ - cantly impact the perceived usefulness ( PU ) . Also, perceived enjoyment and subjective norm exert a signi ﬁ cant impact on the perceived ease of use ( PEOU ) . Additionally, PEOU in ﬂ u - ences the PU, which in turn in ﬂ uences the BI. Furthermore, the ﬁ ndings reveal a strong link between BI and the actual use of the LMS. Finally, the study suggests that students ’ BI to use the LMS is the most important factor for students ’ actual use. PU is also an important determinant factor of students ’ actual use. Following the study ’ s conclusion that BI and PU are important for students ’ actual use of an LMS, higher education administrators must develop rules that increase the students ’ PU of LMSs, while also ensuring that current measures that awaken students ’ BIs are main - tained or improved so that students can use the LMS for academic work.


Introduction
Organizations are devising new business models in order to avoid the ravages of coronavirus. While some organizations allow employees to work from home, others have divided the workforce so that one part can work from home, while the other is physically present in the office. These arrangements are currently available in Ghanaian universities. At the University of Professional Studies, Accra (UPSA), for example, one group is engaged in face-to-face teaching, while the other group relies on distance learning for the first half of the semester. The groups then reverse their mode of study during the second half of the semester. As a result of these arrangements, these universities have invested in learning management systems (LMSs) to manage online learning.
LMS software is a collection of technologies that let educational instructors communicate learning information to students via the internet while also managing and monitoring their learning progress throughout the learning process (Ramesh & Ramanathan, 2013). It provides the "big picture" of learning objectives, content to be taught, learning activities, requisite resources, and class managementall while adhering to the instructional paradigm chosen for the unit (Metzler, 2017). Despite the existence of numerous popular LMSs around the world, such as Sakai, Moodle, Blackboard, ATutor, and others, most institutions in Ghana have installed Moodle's opensource LMS. The UPSA, and the University of Education, Winneba, which are public universities in Ghana, and Methodist University Ghana, a private university, are some of the universities in Ghana using Moodle for online learning. The University of Ghana on the other hand uses Sakai to deliver course content to the learners. Hundreds of millions of students utilised Moodle LMS, a free and open-source LMS. It is extensively utilised in educational institutions because it supports a flexible learning environment, is simple to use on mobile devices, available to everyone, and is incredibly secure.
To avoid disrupting the academic calendar, most universities have resorted to using the LMS to transmit course content to students (Olujuwon, 2021). While the LMS may be effective in certain advanced countries, it may not be so in developing nations. Most of the university remedies have failed to address the problem from the students' standpoint. According to Kaewsaiha and Chanchalor (2019), students find it difficult to use the LMS from outside the university due to issues with internet connectivity, the user interface, using specific LMS capabilities, unclear instructions, confusing evaluation criteria, lack of examples, time to complete activities did not match the actual time required, and receiving no relevant feedback or no feedback at all, among other things. According to Alebaikan and Troudi (2010), higher education institutions must offer instructional practice training for instructors, user-friendly LMS features, emphasize the importance of avoiding plagiarism, and provide orientation, resources, and time management strategies for instructors and students in order to ensure efficiency in online discussions using the LMS.
Many research on the use of LMS by Ghanaian students have employed technology acceptance model (TAM) or a mix of TAM and other factors as their underlying model, as well as a variety of analytical approaches, to investigate the phenomenon. TAM and multiple regression (Essel & Wilson, 2017); TAM and general linear model (Asampana et al., 2017); TAM plus computer self-efficacy, structural equation modelling (Boateng, Mbrokoh, Boateng, Senyo, & Ansong, 2016;Budu, Yinping, & Mireku, 2018); and TAM plus institutional factors and correlation and regression (Okantey & Addo, 2016). The findings of these research show that several variables influence the students' behavioural intentions (BIs) to use and actual use of LMSs. The work of Budu et al. (2018) found perceived ease of use (PEOU), perceived usefulness (PU), and self-efficacy significantly influencing students' BI to use LMS. In a related study by Boateng et al. (2016), they identified attitude and PU as the first and second most significant factors in determining students' BI to use an LMS. Boateng et al. (2016) proposed that it is important to take into account the antecedent of PU and its impact on BI, as PEOU did not directly correlate with the students' BI, but rather, did through PU. Essel and Wilson (2017) argue that the benefits of employing interactive and collaboration technologies to promote learning should be increased because students appear to use these features infrequently. They also proposed that efforts be made to alleviate students' issues with Moodle, such as internet connectivity and navigation, in order to boost their Moodle usage rate.
According to the findings of Adjin-Tettey (2014) from the UPSA, students' perceptions of the Moodle course management system are overwhelmingly positive, and this has a direct impact on users' intentions to use the system for the remainder of their campus studies, as evidenced by the correlation test results. Similar research from the same university, however, by Asampana et al. (2017), indicates that students' acceptance of LMS is very poor, blaming it on a lack of information technology infrastructure, insufficient training, and the system's relevance to great lecture delivery. Furthermore, despite the plethora of studies on students' LMS adoption in the literature, evidence on the motivators of Ghanaian students' BI and actual LMS usage behaviour in respect to the third version of TAM (TAM3) is scarce. Boateng et al. (2016) used a covariance-based structural equation model approach in Ghana, which has less statistical power to examine the students' intentions when several factors are present. In addition, the research did not look at how PEOU and PU were modified by external circumstances. It is therefore important to address the gaps in the literature with regards to Ghanaian students' adoption of LMS.
Therefore, the purpose of this research is to explore the motivators for LMS adoption by UPSA students in Ghana using the partial least squares structural equation model (PLS-SEM) and a simplified version of the TAM3 model. The study specifically attempts to analyse the external factors that influence PEOU and PU, as well as the value of the constructs to students' use of the LMS.

Literature Review
Administrators at higher education institutions deal with risks and uncertainty in a variety of ways. Some are useful, while others have a negative impact on the students' academic achievement. In a period of major curricular changes and vastly variable student performance in a chaotic environment, it is critical to examine and understand the students' behaviour on the use of LMS in relation to their academic achievement.
An LMS is a collection of software tools for managing user learning interventions, courses, instruction, online communication, and assessment and testing (Nof, 2009). It also enriched instruction by encouraging communication among students and between students and teachers, sharing documents by providing access to course materials and assignments, tests and test results, an e-discussion and chat room, and other educational features (Taylor, Handler, FitzPatrick, & Whittingham, 2020). Furthermore, it allows teachers to assess the depth of student learning and provide helpful comments by facilitating self-directed learning practice (Dent, Harden, & Hunt, 2021). The most extensively used LMS is Moodle in the Ghanaian context.
Despite the LMS's many features, there are several drawbacks to its success when it comes to student usage. According to Ssekakubo, Suleman, and Marsden (2011), the most common causes of LMS failure are poor levels of student familiarity with technology, low levels of information and communication technology literacy, and a lack of assistance. The hexagonal e-learning assessment model identifies six types of success factors that influence e-learner satisfaction (supportive issues; social issues: learner perspective; social issues: instructor attitudes; technical issues: system quality, information (content) quality, service). Notwithstanding the significant expenditure to build and maintain the LMS and other technologies, there is evidence that users do not fully utilise them (Mtebe, 2015).
The interaction between students and LMS during the COVID-19 period in Ghana has been the subject of several research, which have revealed significant evidence. Dampson et al. (2020) contend that most users of the University of Education, Winneba will stop using the LMS platform for teaching and learning activities after COVID-19, and that the use of the LMS will cease whenever users have the option of choosing between the LMS and the traditional classroom approach to instruction. The majority of University of Cape Coast (UCC) students, according to Agormedah et al. (2020), were aware of e-learning and use tools like Google Classroom, Alison, and the UCC Moodle platform. However, they were unprepared for the transition to remote learning because they had not been introduced to it, and they perceived constant internet connectivity as a challenge due to financial insecurity, which leads to resistance from some students, which could affect their remote instruction engagement activities and academic success. Furthermore, Agormedah et al. (2020) argue that students' unpreparedness, as well as a lack of formal training and experience in e-learning platforms, can have a negative impact on their behaviour outcomes such as engagement in learning, satisfaction, participation, motivation for learning, online work skills proficiency, self-directed learning, and efficacy in the use of e-learning devices, as well as their academic performance. In a related study in Ghana, Opoku (2021) found that learning distractions from family, environment, and technology negatively impact students' Moodle usage, which in turn negatively impacts the academic performance. The study further reveals that students' academic performance had declined after taking tests and quizzes during the Covid-19 lockout.
TAM was created to forecast information technology acceptance and utilisation on the job. It is suitable for information systems environments (Lee, Kozar, & Larsen, 2003). TAM has been applied to a wide range of technologies in a variety of settings. According to the authors, PU and PEOU have a direct impact on BI. In addition, the authors claimed that PEOU has a direct impact on PU. Venkatesh and Davis (2000) examined the consistent and large influence of PU on BI across all research investigations, as well as the inconsistent effect of PEOU on BI. Venkatesh and Davis (2000) devised the extended TAM (TAM2) as a way of studying the elements that influence PEOU after discovering these trends in the literature. TAM2 incorporates additional social influence processes (subjective norm (SN), voluntariness, and image) as well as cognitive instrumental processes (job relevance (REL), output quality (OUT), result demonstrability, and PEOU). Also, Venkatesh (2000) looked into the factors that influence PEOU and came to the conclusion that individuals form early impressions of a system's PEOU based on a number of factors related to people's general attitudes toward computers and how they are utilised. Venkatesh and Bala (2008) established TAM3 by integrating the factors from Venkatesh and Davis (2000) and Venkatesh (2000). Figure 1, which was adapted from Venkatesh and Bala (2008), shows a graphic representation of TAM3. TAM3 expands on the technology acceptance model by arguing that image, REL, OUT, result demonstrability, and SN all predict PU. Furthermore, they propose that computer self-efficacy, perception of external control, computer anxiety, and computer playfulness are anchor variables that have a direct influence on PEOU. The model also suggests that experienced enjoyment and objective usability are adjustment factors that predict PEOU. According to the authors, experience moderates the relationships between (i) PEOU and PU, (ii) computer anxiety and PEOU, and (iii) PEOU and behavioural intention. The relationship between SN and BI is also moderated by voluntariness of use.
Although TAM3 could not account for crossover effects, Abdullah and Ward (2016) developed the General Extended Technology Acceptance Model for E-Learning to be used in the context of e-learning adoption.
Recent empirical studies on technology adoption in higher education that used TAM or a modified version of TAM as a conceptual framework to investigate students' adoption of technology suggest that PU and PEOU have varying effects on students' intention to use the technology. A slew of research has confirmed that PU and PEOU have a major impact on students' intentions to use LMS  discovered REL as an important variable in determining students' LMS PU. Furthermore, Hanif, Jamal, and Imran (2018), as well as Abdullah, Ward, and Ahmed (2016), demonstrated that students' perceived enjoyment (ENJ) determines PEOU of LMS. Furthermore, Agudo-Peregrina et al. (2014) confirmed that personal innovation in IT has an impact on PEOU in higher education. In light of the disparate findings from the preceding studies, which may be attributed to differences in culture and student familiarity or experience with LMS, a study employing a modified TAM to explore the factors affecting UPSA students' LMS use behaviour (UB) is required.

Conceptual Framework and Hypotheses
The technology acceptance model and its extensions have been applied in a variety of situations and have been found to be useful in explaining students' intention and actual use of LMS (Al-Azawei et al., 2017; Binyamin et al., 2018;Kaewsaiha & Chanchalor, 2020). Thus, the study adapted TAM3 as the base model for the study. There were four core variables: REL and OUT, which directly influence PU, and SN and ENJ, which directly influence PEOU. There are three mediating variables (PEOU, PU, and BI). Finally, actual UB was an endogenous variable. The conceptual model that the study is based on is depicted in Figure 2.

ENJ
ENJ is the degree to which "the activity of using a specific system is perceived to be enjoyable in its own right, apart from any performance consequences resulting from system use" (Venkatesh, 2000, p. 351). The authors of TAM3 contend that ENJ has a direct influence on PEOU. Al-Gahtani (2016) validated this assertion in a higher education setting. As a result, the current study posits that: H1: ENJ significantly influences students' PEOU of LMS.

SN
The extent to which a person believes that the majority of influential people believe he should or should not use the system (Venkatesh & Davis, 2000). These important people can be a friend, family members, or people higher in authority. Although the original TAM2 posits a direct and indirect relationship between SN and PU, other studies (Farahat, 2012) have also shown that SN relates to PEOU. Thus, this study posits that: H2: SN has a significant impact on PEOU of LMS.

REL
REL is defined as an individual's perception regarding the degree to which the target system is applicable to his or her job (Venkatesh & Davis, 2000). According to Venkatesh and Davis (2000) research, there is a strong relationship between REL and PU. In this instance, the study defined REL as the extent to which a student considers that the use of LMS is suitable for learning and performing learning-related task ( Thus, the study posits: H3: There is a significant relationship between REL and PU.

OUT
OUT is the degree to which an individual believes that the system performs his or her job tasks well (Venkatesh & Davis, 2000). The extended model of TAM (Venkatesh & Bala, 2008) also substantiated the relationship between OUT and PU. Students are also deemed to perceive the LMS as useful when they are able to perform academic work satisfactorily. Previous research (Teo, Zhou, Fan, & Huang, 2019) validated the relationship between OUT and PU in a higher educational context. Thus, the study posits that: H4: OUT has a positive effect on PU.

PEOU
PEOU refers to the extent to which a person considers that the use of a system is free of effort (Davis, 1989). It also refers to the difficulty of using an LMS as a tool for accessing course materials. The PEOU includes the following items: physical and mental effort needed; understandability of the use cases; ease of learning for operating various usages; operational efficiency of the use case in terms of error-proneness, controllability, unexpected behaviour; and user-friendliness in terms of ease of remembering and guidance (Grover, Kar, Janssen, & Ilavarasan, 2019). According to TAM, PEOU influences both BI and PU (Venkatesh & Davis, 2000). Prior empirical studies (Al-Gahtani, 2016; Kaewsaiha & Chanchalor, 2021) in the higher educational context support these views. Hence, the study posits that: H5. PEOU predicts the PU of the LMS by students. H6. PEOU positively predicts students' BI to use LMS. Davis (1989) referred to PU as the extent to which a person believes that a system may contribute to improving their work performance. In the educational context, it may be redefined as the extent to which a student believes that the LMS may help to improve one's academic performance, by enhancing the learning process through easy access to learning materials and by having interactive sections with classmates and lecturers. Davis (1989) averred a significant relationship between PU and BI to use an innovation. Other works (Al-Gahtani, 2016; Kaewsaiha & Chanchalor, 2021) in higher educational institutions have validated this assertion. Thus, this study posits that: H7: The PU of an LMS has a positive effect on the BI.

BI
BIs are indications of a person's readiness to perform a behaviour (Fishbein & Ajzen, 2011). In other words, BI is the person's estimate of the likelihood or perceived probability of performing a given behaviour. According to Fishbein and Ajzen (2011), the readiness to act, symbolized by an intention, is often expressed in statements like these: I will engage in the behaviour, I intend to engage in the behaviour, I expect to engage in the behaviour, I plan to engage in the behaviour, and I will try to engage in the behaviour. Davis (1989) posits a direct relationship between BI and actual usage. Other empirical studies (Binyamin et al., 2018;Kaewsaiha & Chanchalor, 2021) have supported this claim in the context of LMS use in higher education. Therefore, the study postulates that: H8: Students' BI to use LMS significantly influences their actual UB.

Methodology
The study used cross-sectional survey design and a convenient sampling technique to elicit data from the population of first-year students in the faculty of information technology and communication studies at UPSA to address the objective of the study. The study chose that population because they were new to using the LMS for academic work. The findings may drive administrators to develop rules that could encourage new students to use the LMS. According to Hair, Hult, Ringle, and Sarstedt (2017), PLS-SEM performs well with fewer data and a more complicated model and makes few assumptions about the underlying data. Thus, due to the study's structural model complexity, which includes various variables, indicators, and model linkages, the study used PLS-SEM to better understand UPSA students' behaviour by exploring TAM3 performance in the Ghanaian educational context. The study relied on first-year student class representatives from the faculty to identify volunteers who agreed to participate in the survey. Following that, they received an email at the end of the academic year where all students have used the LMS with a link to a Google form. Only 188 of the 200 volunteers completed the survey.
The online survey was created with Google Forms by adapting Venkatesh's and Bala's (2008) questionnaire items and scales. The items were assessed using a sevenpoint Likert scale. As part of the survey's introduction, participants were told that the responses they supplied would be kept private. The second and third sections were utilised to collect demographic information and data relevant to the study's goal. Prior to collecting data, the Institutional Research Ethics Committee granted its approval for this study. All individuals provided informed consent prior to participation in the study, which was carried out in compliance with the ethical guidelines specified in the Belmont Report and the Declaration of Helsinki. The rights, welfare, and dignity of the study participants were protected by the researchers according to all ethical rules and regulations established by the institutional review board or ethical review committee To evaluate the reflective measurement model, the study used composite reliability and individual indicators to examine the internal consistency reliability, the average variance extracted (AVE) to test convergent validity, and the Heterotrait-Monotrait ratio (HTMT) to assess discriminant validity. Composite reliability values range from 0 to 1, with higher values indicating higher levels of reliability. In exploratory research, measured values of 0.600-0.700 as composite reliability are acceptable. In complex research, values between 0.700 and 0.900 are considered satisfactory (Hair et al., 2017). Furthermore, factor loadings greater than 0.708 and AVE greater than 0.500 are sufficient to assess convergent validity. According to Hair et al. (2017), to confirm discriminant validity, the HTMT values must be less than 0.850 for the stricter condition and less than 0.900 for the lenient condition; additionally, the HTMT values must be significantly different from 1.
To evaluate the structural model, the study looks at the model's predictive capabilities as well as the relationships between the latent variables. The primary evaluation criteria for PLS-SEM results are the check for multicollinearity (VIP), coefficients of determination (R 2 ) values, and the t-values of the path coefficients. The study used the bootstrapping routine to assess the significance of the path coefficient. As a rule of thumb, the study used 5,000 subsamples, which exceeded the number of valid observations. Other measures examined in the study include effect size (f 2 ), predictive relevance (Q 2 ), and effect size (q 2 ). Path coefficients in structural models represent the hypothesized relationships between latent variables. Critical values for a two-tailed test at 10, 5, and 1% significance levels are 1.65, 1.96, and 2.57, respectively. In this study, path coefficients with a 5% chance of error were considered statistically significant.

Profile of Respondents
The respondents for this study are first-year students who have only used the LMS for one academic year. The distribution of the respondents suggests that 44.1% were females and 55.9% were males. The majority of students were between the ages of 18 and 24, accounting for 80.3% of the total, with the next highest percentage being between the ages of 25 and 30, accounting for 18.1%. Students above 30 years constitute 1.6%.

Measurement Model Appraisal
The results of the measurement model appraisal shown in Table 1 indicate that the item loadings range from 0.652 to 0.904 and the AVE ranges between 0.509 and 0.752, which exceeds the minimum threshold indicating that the construct explains at least 50% of its items' variance. Most researchers use composite reliability to assess internal consistency reliability, but Cronbach's Alpha is another measure of internal consistency reliability. While Cronbach alpha is too conservative, composite reliability is also too liberal and the construct's reliability is viewed within these extreme ends. The rho_A values presented in Table 1 are approximately an exact measure of construct reliability.
The study assessed the discriminant validity using the HTMT criterion (Hair et al., 2017) and the confidence interval bias was corrected.  Table 3 demonstrates that the columns labelled 2.5 and 97.5% show the lower and upper bound 95% confidence interval bias corrected, and none of the confidence interval values include 1, indicating that the constructs are distinct. As a result, the reliability and validity of the measures are supported by all model evaluation criteria.

Structural Modelling
The results of each predictor construct's tolerance, variance inflation factor (VIF), given in Table 4 fall within the acceptable range of 0.2-0.5 using the rules of thumb for structural model evaluation outlined by Hair et al. (2017). As a result, there are no issues with collinearity between the constructs.
To determine the significance of path coefficients, the researcher utilised a bootstrapping approach with 5000 samples. Apart from the association between PEOU and BI, which is negligible, the bootstrapping findings in Figure 3 and Table 5 reveal that all the relationships in the model are significant at the 5% significant level. As a result, the data support hypothesis H1, H2, H3, H4, H5, H6, and H8, but not H7.
The coefficient of determination (R 2 ) shows the amount of variance in the endogenous constructs explained by all the exogenous constructs linked to it (Hair et al., 2017). That is, the coefficient measures the amount of variance in the endogenous constructs explained by all of the exogenous constructs associated with it. It is thus a measure of the model's predictive power. Generally, R 2 values of 0.250, 0.500, and 0.750 are considered to be weak, medium, and substantial, respectively (Hair et al., 2017). Thus, the R 2 values shown in Table 6 suggest that ENJ and SN explained  The effect size is used to determine how much a predictor construct contributes to the R 2 value of the target constructs (f 2 ) (Hair et al., 2017). Effect sizes of 0.02, 0.15, and 0.35 are evaluated as small, medium, and substantial, respectively, by Hair et al. (2017). The f 2 values are shown in Table 6. While SNs have a moderate influence on PEOU, enjoyment has a substantial effect. PEOU and REL have medium impact sizes on PU, but OUT has a modest effect size. Furthermore, in terms of BI, PU has a large effect size, but PEOU has a relatively tiny effect size. Finally, the impact of BI on actual UB is significant. Table 6 further shows the predictive relevance of the model with respect to endogenous variables. Since Q 2 is greater than 0, the findings clearly corroborate the model's predictive relevance with regard to the latent variables. The impact sizes of ENJ and SN are medium and small. OUT and REL also have a small impact size on PU. Furthermore, PU and PEOU have medium and small impact sizes on BI.

Importance-Performance Map
Analysis (IPMA) The IPMA seeks to identify which elements have poor performance but high importance for the target construct   (Hair et al., 2017). The IMPA is a reliable and practical analysis in PLS-SEM that extends the usual path coefficients in a more practical method. In this study, BI to use the LMS predicts the students' actual use of the LMS. The precursors of BI are PU and PEOU. REL and OUT, enjoyment and SN, respectively, predict the PU and PEOU. The study employed UB as the target construct and ran the IPMA. The performance scores were created by rescaling the latent variables score range, with 0 being the lowest score and 100 being the highest. The importance scores were obtained from the total direct influence of outcome variables using the structural equation model (Hair et al., 2017). Table 7 presents the total effects (importance) and index values (performance) used for the IPMA. It can be seen that BI is the most important factor in order to determine the UB of LMS by students due to higher importance values compared to other latent variables. PU is at an intermediate level, while PEOU has the lowest importance level after PU. The level of importance and performance are shown in Table 7.
According to the IPMA in Figure 4, SN and REL are in the overkill quadrant, ENJ, OUT, and PEOU are in the lower priority quadrant, PU is in the concentrate here quadrant, and BI is in the keep the good work quadrant (Martilla & James, 1977).
The IPMA in Figure 4 clearly shows that higher education institutions must take suitable efforts to increase the students' PU of LMS. This can be accomplished by utilizing resources that can be used to enhance students interaction among themselves and the relevance of LMS's on enhancing students' PU of LMS. Higher education administrators must also continue the positive effort that inspires students' BI to use the LMS.

Discussion
The LMS is widely used by higher education institutions that use either an entirely online or a blended learning approach. However, only a few studies have used an extended TAM to investigate the determinants of Ghanaian higher education students' BI and actual use of LMS. Based on the TAM3, this study chose two external variables to predict PEOU (SN and ENJ) and PU (REL and OUT). The results showed that the predictors of PEOU and PU explained 60.1 and 61.9% of the variance, respectively. Furthermore, the model explained 52.9% of the variation in BI and 45.6% of the variation in actual UB, respectively. Although the effect size of SN on PEOU is moderate, the effect size of ENJ is significant, and the effects of PEOU and REL on PU are moderate whereas the effect size of OUT is small. Given ENJ's significant predictive power on PEOU and SNs' moderate predictive power on PEOU, as well as REL's and PEOU's moderate predictive power on PU, it is crucial to pay attention to these factors while using LMS as a study tool.
First, the results show that ENJ is positively related to PEOU of LMS. This demonstrates that as students have fun with the use of the LMS, students' perception of LMS' ease of use increases. This finding is consistent with the findings of Venkatesh and Bala (2008) and other studies in higher education institutions (Al-Gahtani, 2016;Hanif et al., 2018). The students' attention is drawn away from any difficulties in using the LMS due to their happiness as a result of their use of the LMS. It is therefore critical for designers to ensure that the LMS interface is user-friendly. Furthermore, essential options must be easily accessible to students in order for them to complete their assignments.
Second, the research found a connection between SN and PEOU. This illustrates that as key persons in students' lives encourage them to use the LMS, students perceive the LMS to be simple, which increases the BI to use the LMS, which effects actual usage. This result is in line with the previous work (Farahat, 2012;Revythi & Tselios, 2019) in a higher education context. The finding, however, contradicts previous research in a similar context (Park, Nam, & Cha, 2012;Salloum, Alhamad, Al-Emran, Monem, & Shaalan, 2019;Binyamin et al., 2018). Higher education administrators must devise ways for peers and tutors to make other students aware of the LMS's value.
Third, the study showed a strong positive relationship between REL and LMS PU. According to the findings, the more students perceived the LMS to be relevant to them, the more they perceived it to be useful. The outcome is consistent with the work of Kaewsaiha and Chanchalor (2020) and Venkatesh and Bala (2008) in the higher education context. As a result, it is critical for LMS administrators to ensure that students realise the significance of the LMS to their studies by responding to all academic concerns about its use.
Fourth, the study established that there is a significant positive relationship between OUT and PU of LMS, resulting in the higher the quality of the LMS output, the more students regard the LMS as useful. This finding supports the findings of Venkatesh and Davis (2000) and Venkatesh and Bala (2008). The finding is also consistent with the findings of Teo et al. (2019) in higher education settings. When the course materials on the LMS are simple, adequate, and easy to understand, students perceive OUT to be high. They also believe that lecture delivery is comparable to face-to-face learning and that internet connectivity is stable. Higher education administrators must ensure that course materials are available online. Fifth, the study discovered that PU is influenced by PEOU. This finding suggests that as long as LMS are simple to use, students see them as useful. This conclusion is similar to that reached by Venkatesh and Bala (2008), as well as other empirical studies in higher education (Al-Gahtani, 2016; Kaewsaiha & Chanchalor, 2021). In a comparable setting, however, this result contradicts Saroia and Gao (2019). Though the relationship between PEOU and BI is small, it is consistent with the findings of Agudo-Peregrina et al. (2014), Boateng et al. (2016), and Binyamin et al. (2018) in the educational setting. It should be noted that there is an indirect relationship between PEOU and students' BIs. Therefore, it is critical   for administrators and designers to ensure that students perceive LMS use to be simple for learning. Sixth, the study revealed that PU has a significant effect on BI. This finding shows that as students perceive the LMS to be useful, their proclivity to form an intention to use the LMS increases. This relationship supports the claims made by Davis (1989) and Venkatesh and Bala (2008). The discovery is also consistent with previous empirical work in higher education (Agudo-Peregrina et al., 2014;Al-Gahtani, 2016;Kaewsaiha & Chanchalor, 2021). The IMPA results also clearly show that PU is important to students' actual use of the LMS. As a result, resources that may be utilised to enhance SNs and REL but are overkill can be used to improve students' PU of LMS. Higher education administrators must develop rules that enhance students' PU of LMSs.
Finally, the findings show that BI and actual usage have a substantial positive association. This finding shows that when students' BI to use a LMS grows, so does their actual use of the system. Venkatesh and Bala's (2008) hypothesis is supported by this finding. The result is also consistent with other empirical investigations in the field of higher education (Binyamin et al., 2018;Kaewsaiha & Chanchalor, 2021). However, the findings contrast those of Agudo-Peregrina et al. (2014). As a result, regardless of how simple or complicated the LMS is to use, administrators must ensure that learners find it useful. The IMPA finding confirms the importance of BI in students' LMS use. As a result, higher education administrators must ensure that policies set in place to awaken students' BIs are either maintained or improved so that students can use the LMS for heightened academic work.

Conclusion, Limitation, and Future Direction
The current study sought to explore the motivators for LMS adoption by UPSA students in Ghana using the PLS-SEM and a simplified version of the TAM3. The study evaluated the key factors that influence LMS usage and offers some implications for stakeholders in the higher education sector. The study drew a sample of 188 students from the faculty of Information Technology of the UPSA in Ghana. A modified version of TAM3 is used as the study's framework. According to the study's findings, BI has a significant influence and importance on students' actual use of LMS. Though there is no direct relationship between PEOU and BI, there is an indirect relationship between PEOU and BI. Finally, REL and OUT had a significant impact on PU, whereas SN and ENJ had a significant impact on PEOU.
The findings of the study have a wide range of implications for practitioners, policymakers, and researchers. Given that ENJ has a significant influence on PEOU, LMS designers must ensure that students enjoy using it. This can be accomplished by ensuring that the LMS interface is consistent with other systems, easy to use, and standards-compliant. Administrators must ensure that students have easy access to course materials and that they can be used without difficulty. Policymakers can also collaborate with internet service providers to keep data costs low and internet access available in every corner of the country.
After determining that a SN predicts PEOU, higher education managers must persuade tutors to advertise the LMSs' worth to students in order to encourage them to use it. Administrators must also guarantee that course materials available on the LMS are up to date and relevant to students' academic work, given that REL predicts PU. Again, administrators must ensure that the LMS contains all the learning resources required by students for the course, having determined that OUT is connected to PU. Finally, lecturers must respond to student complaints on time and comment or mark all assignments distributed to students via the LMSs.
Though the study identified some important relationships that could aid in the resolution of issues related to the use of LMS, it is also important to acknowledge some limitations associated with this study. Despite the fact that external variables explained 61.9% of PU and 60.1% of PEOU in TAM3, the study did not include all of the external variables used in TAM3. Future studies could make use of all TAM3 variables as well as other additions. As in the original TAM3, the study failed to address the moderating effect of experience and voluntariness. Future research could look into the moderating effect of these variables. This study's participants were drawn solely from a public university. When generalizing the findings to include all university students, caution must be exercised. In the future, efforts may be made to reach all public, private, and technical university students in the country.
Acknowledgments: To everyone who helped with the effective completion of this piece and its publishing, I would like to offer my profound gratitude. I want to express my gratitude to my institution, UPSA and students for their invaluable support with the data collection and processing. Their meticulousness and focus on the little things helped to guarantee the excellence and correctness of our findings. My sincere gratitude goes out to the editors and reviewers who helped shape this work by offering insightful comments and useful criticism as the article was being developed.
Funding information: This research received no specific grant from any funding agency, commercial or nonprofit sectors.