Evaluating the effects and outcome of technological innovation on a web-based e-learning system

Abstract The advancements in technological innovations have had a substantial impact on the Internet infrastructure, Web technology, and the usefulness of e-learning systems. This requires a discrete initiative to exploit the full capabilities of technological innovation to improve the functional characteristics of e-learning systems, the ensuing benefits, and to sustain educational challenges. A context-based model that incorporates explicit and widely accepted determinants of information systems success, technology acceptance, e-learning success models, and technological innovation has been developed. Empirically, the model has been validated using a logistic regression method and 458 valid responses. Findings indicate a unit increase in the level of technological innovation results in 55 times improvement in the features of e-learning systems and 3 times in e-learning outcome. Also, a unit increase in the features of e-learning systems results in three times improvement in the outcome of e-learning system. Risk minimization and achieving learning goals are the most important positive benefits of e-learning systems to students. The factor with the highest positive impact on e-learning systems is improved software service. 52% of students indicated e-learning systems are innovative. 61% specified that the usefulness and performance of e-learning system(s) are satisfactory and would strongly recommend them to others.


PUBLIC INTEREST STATEMENT
In this pandemic era of COVID-19, the educational systems globally are affected making webbased electronic learning systems learning preferences to sustain educational challenges. This requires a discrete initiative to exploit and provide unique insights on significant factors to improve its usefulness. This study identified and evaluated specific factors that impact the usefulness of an e-learning system. Using survey data from 458 participants, relationships among technological innovation, quality factors of information systems, technology acceptance factors, user beliefs, and e-learning benefits were assessed. Our findings show that a unit improvement in the levels of technological innovation results in 55 times improvement in the features of e-learning systems which in turn improves the outcome of e-learning 3 times. Risk minimization and achieving learning goals were the most important positive benefits of e-learning systems to students. 61% of students were satisfied with the usefulness, and would strongly recommend e-learning system(s) to others.

Introduction
Technology advancements have had a profound impact on the learning and teaching activities of education. Several learners have embraced technology-driven educational activities (Alkandari, 2015). This has necessitated employing the most suitable and advanced technology for effective e-learning delivery, engaging learners, boosting learners' motivation, and improving satisfaction and learning productivity (Kangas et al., 2017). As a technology-mediated learning management system, e-learning is considered an essential learning medium in the higher educational sector to support educational services (Al-Fraihat et al., 2017).
A significant volume of research has been conducted to identify, examine, and evaluate important factors that influence an e-learning system with the intent of advancing and maximizing its usefulness (Fathema et al., 2015;Mtebe & Raphael, 2018). Eom and Ashill (2018) are of the view that developing a comprehensive model that cuts across multiple levels to understand synergistic effects of the key determinants of a successful e-learning system is imperative. However, researchers face the challenge of dealing with an excessive number of important factors that determine the success of an e-learning system in literature (Al-Fraihat et al., 2020). Due to the fact that the core factors that determine the success of an e-learning system differ in their importance in a specific context, a research that focuses on exploiting the dynamic capabilities of technological innovation to advance the functional characteristics of e-learning systems, ensuing benefits, and to sustain educational challenges is necessary. Technological innovation is making an e-learning system a learning preference since it facilitates the development of value-added and interactive features, components, services, and methods of accessing an e-learning system.
To help create an effective and innovative teaching and learning environment, this study proposes a context-based model that incorporates explicit factors of widely accepted models and specific factors of technological innovation that are of recent concern in the e-learning domain.

A web-based e-learning system
The Web as a techno-social system is the largest transformable-information construct. The Web is the most prominent part of the Internet and provides one of the most popular services of the Internet. Enhanced content, ease of use, availability, attractive design, offering value proposition, using analytical tools, regular updates are among the critical success factors of a website. An ease of use of a web's interfaces is a quality attribute, hence, an essential condition for a useful Website (Nielsen, 2012). The major quality attributes include ease to learn, effectiveness, ease to recall, minimal errors made, fulfilment, and utility (design's functionality) (Nielsen, 2012). Presently, the Internet is promoting self-paced learning across the globe, and has made it easy to obtain different formats of learning resources, to teach and regulate courses, and to expand interactive communication and collaboration tools (Al-Fraihat et al., 2020). An e-learning system could be virtual-based, web-based, computer-based, or digital-based (Abdellatief et al., 2011). Generally, authors have proposed technology-driven, delivery-system-oriented, communication-oriented and educational paradigm as the main types of e-learning elements (Arkorful & Abaidoo, 2015). Some popular web-based e-learning systems available are WebBoard, Blackboard, Glow, Google classroom, Code academy, MOODLE, and Sakai.
The evolution of the Web and continuous advancements in technology have resulted in several definitions of e-learning. E-learning is defined by Choudhury & Pattnaik as "the transfer of knowledge and skills, in a well-designed course content that has established accreditations, through an electronic media like the Internet, Web 4.0, intranets, and extranets." (Choudhury & Pattnaik, 2020, p. 2). An e-learning system is a type of information system (IS) that supports improving technical system qualities including dependability, accessibility, and usability of a system (Al-Fraihat et al., 2020), service quality enhancement through technology interface (Parasuraman, 2010), and improvement of information quality (Anderson et al., 2014). Quality features influence the processing and delivery of information and the adoption of e-learning as a learning tool (Adel, 2017). Choudhury and Pattnaik (2020) reviewed 138 articles published between 2000 and 2018 and identified certain critical success factors of e-learning that pertain to various e-learning stakeholders. Some of these critical success factors include updated technology, appropriate course and interface design, social presence, computer literacy, technology, and application interoperability, course customization, interactivity, ease of use and autonomy, collaboration among stakeholders, and motivation. From this same research, Choudhury and Pattnaik (2020) identified lack of management' support, and continuous innovations and fast technology advancement (for instance, matching state-of-the-art designs and technology) are as some challenges in e-learning.

Integrating web with e-learning
Web tools have been used progressively in educational contexts. Web 1.0 to web 5.0 has really transformed the delivery of learning electronically. According to Wentling et al. (2000), Internetbased learning was introduced with Web 1.0 where knowledge acquisition and distribution were essential in the year 2000. As at 2016, complete degree programmes were delivered online and remote interaction with an instructor was possible as at 2018 (Ali et al., 2018). Web 2.0 is a readwrite web that facilitates designing new models of e-learning and creation of knowledge systems (Aghaei et al., 2012). For instance, developers use HTML/CSS, JavaScript, and PHP to develop the dynamic features of a web-based e-learning system to address new trends of adaptation. With this, characteristics of learners and their learning styles can be understood to generate personalized and adaptive user interfaces and recommend course contents (Kolekar et al., 2018). Flex, Asynchronous JavaScript and XML are some of the basic development approaches that developers use to create the applications of web 2.0.
For Web 3.0 (Semantic Web), the core software technology is intelligently learning and understanding semantics where service and process quality such as search personalization is improved (Vieira & Isaías, 2015). For Web 4.0 and Web 5.0, there is synergy between humans and machines, emotional dimensions are added to enhance interactions, and are considered autonomous (Parvathia & Mariselvi, 2017). Since e-learning is a technology-mediated learning process, and web technology has had a profound impact on the learning and teaching methods of education, successful development, and implementation of an e-learning system as well as delivery of services mainly depend on innovation activities.

Technological innovation support for e-learning
Technological innovation is a significant source of sustainability (Eidizadeh et al., 2017) and a longterm business strategy (Su & Tang, 2016) that thrives on the capabilities of information technology (Bassellier & Benbasat, 2004). Rapid diffusion of IT drives a technological innovation effort to promote information dissemination, firm-wide networking, collaboration, and improvements in communication (OECD, 2010). Innovation capabilities are unique and valuable resources of a firm that leverage structural resource differences to achieve quality among competing firms (Benitez-Amado & Walczuch, 2012). Technological innovation facilitates the development of features, components, services, and methods of accessing an e-learning system (Obeng & Boachie, 2018).
Process and product innovations constitute technological innovation. A new or significantly enhanced features of goods and/or services is referred to as product innovation (OECD, 2005). Thus, technical specifications, software components, usability, and other features are improved. Primarily, work on product innovation is effectiveness-driven that addresses market needs (Bergfors & Larsson, 2009). A new or significantly enhanced technique, equipment and/or software for production or service delivery is referred to as process innovation (OECD, 2005). As a distinct initiative, process innovation requires change management of core organizational operations. Through process innovation, services are automated. Ding and Straub (2008) are of the view that, in a fully automated IT-service delivery context (see Figure 1), services are delivered through IT artefacts electronically (online) with minimal human involvement. In a web-based e-learning system, the service channel between the learner and service provider is considered an information technology artefact. In such a self-service context, quality of system and information are offered by an IT artefact that are considered integral parts of service quality (Ding & Straub, 2008) and influence the learner's perception of service quality. Practically, IT facilitates quality user-interface design that exhibits good screen layouts and explicit instructional support for a self-paced e-learning tool that makes users comfortable to use the tool (Liu et al., 2010).
Technological innovation has positive impact on perceived usefulness, intent to use (Ngafeeson & Sun, 2015), perceived satisfaction (Joo et al., 2014), and the adoption of electronic learning technology. Continuous developments in Internet infrastructure, innovations, and the World Wide Web technology have made electronic learning systems more flexible, usable, interactive and a learning preference (Alkandari, 2015;Wang et al., 2019). Using a context-based model that incorporates explicit factors of technology acceptance and adoption, information systems success, and user satisfaction models, the study focuses on identifying and evaluating specific factors of technological innovation that impacts a web-based electronic learning system.

Conceptual model development
Main approaches, perspectives, and measurements that had been used to evaluate the success of IS and e-learning were adopted to develop the proposed conceptual model to assess the effect of technological innovation on a specific context of an e-learning system. A web-based e-learning system is a type of information system. Several models have been created by researchers purposely to emphasize the need for appropriate and more consistent success metrics and to explain what makes some IS successful (Delone & McLean, 2003;Petter et al., 2008). According to DeLone and McLean (1992), Delone & McLean (2003)), models help determine, understand, and provide sparing explanation if there is existence of causal relationships among dimensions of success. For Davis (1989), models are important for their theoretical and practical values by providing better measures to predict and explain system use. We adopted models to use tested and proven measures, able to compare and validate findings, and to contribute to further development and validation of the measures.
The following themes were used to develop the model ( Figure 2): • Technological innovation (product and process innovation) • Electronic learning system ○ system quality, service quality, information quality ○ perceived satisfaction, perceived usefulness ○ acceptance/usage • Outcome of using an electronic learning system (benefits) A context model was necessary to identify and evaluate specific factors that contribute to the usefulness of an electronic learning system. Since an electronic learning system is an information system and technology, quality factors would describe its features, user benefit would determine usefulness and satisfaction perceived from such technology, and acceptance will deal with usage of that technology. Constructs/predictors and determinants are found in Appendix A.

Constructs adopted from existing models
Using IS success model to assess the usefulness of an electronic learning system is well accepted among researchers (Al-Fraihat et al., 2020). However, there are conflicting findings among some studies. For instance, in 2007, Lin conducted a study that showed a significant effect of quality factors on the actual use of an online learning system, while in 2018, a study conducted by Cidral et al found an insignificant association among quality dimensions and use. According to Eom (2015), improving the explanatory power of IS success model requires the understanding of quality factors of an electronic learning system, hence, a further research is necessary. The constructs of DeLone and McLean IS success model (Delone & McLean, 2003) and other constructs/factors from widely accepted models and theories were adopted for the study to reflect the context of electronic learning. An ease of use, ease to learn, response time, understandability, availability, and reliability are considered some essential quality factors of a system. Information quality is concerned with content design quality, usability, completeness, accuracy, relevance, and timeliness while service quality focuses on help provision, accuracy, reliability, flexibility, responsiveness, and accessibility.
We embraced perceived usefulness and perceived satisfaction components of Technology Acceptance Model (TAM) into our model. We did so because TAM3 focuses on the determinants that influence perceived usefulness and perceived satisfaction of a given innovation (Venkatesh et al., 2017). Acceptance, in terms of actual system use was added to our model by operationalizing it as a factor that determines a successful electronic learning system. Benefits was added as a factor that determines the outcome of using electronic learning systems (Abdullah & Ward, 2016;Seddon, 1997). Other important constructs developed by OECD (2005) (2020) were included in this study.

Purpose of study
Core factors that determine the success of an e-learning system differ in their importance in a specific context. Purposely, this study focuses on exploiting the dynamic capabilities of technological innovation to advance the functional characteristics of e-learning systems and ensuing benefits by developing and testing a proposed context-based model. In an attempt to shape this purpose of study, the researchers sought answers to questions below and used the hypotheses that follow to drive these research questions.
(1) How does technological innovation impact e-learning systems?
(3) What important benefit(s) do the users of e-learning systems receive?

Methods
The study followed a causal-comparative quantitative approach to test the theoretical model and related hypotheses, establish cause-effect relationships among variables, understand trends in the data, and generalized results after comparing findings with past studies (Creswell, 2012). Crosssectional survey approach was used to collect reliable and accurate data quickly.

Collection of data and preparation
Empirically proven and measured items used to develop the questionnaire were obtained from literature reviews of e-learning, information systems success, and technology acceptance. As a complementary step, and to affirm measurement items of the study, we solicited opinions of persons and peers who are knowledgeable in e-learning (Walker & Fraser, 2005). These experts were tasked to assess whether each item on the 3-point scaled questionnaire was vital, significant (but not vital), and insignificant. Sections to solicit additional inputs from the experts were provided. Responses from the experts showed either the item was essential or important. The results of the Cronbach's alpha coefficient test of variables after incorporating the views of the experts was.77. With this, the reliability of the measurements was confirmed.
Purposive sampling approach was used to select participants studying at various tertiary educational institutes in Ghana since they can provide the necessary information to the study. The finetuned questionnaires were administered using both personal (face-to-face) and online approaches. Trained research assistants were engaged to administer the questionnaire between September and November, 2019. For the data collection in April 2020, a web-based questionnaire was used to minimize the risk associated with printed documents due to COVID-19.
The first part of the questionnaire constitutes the respondents' data and their general views of electronic learning systems. Technological innovation, electronic learning system, and the benefits of using an e-learning system were captured at the second, third, and last sections, respectively. A total of 473 responses out of 600 questionnaires were received (online 291 out of 418, personal 182). A response rate was 76.3% based on 458 responses that were considered valid for further analysis. The characteristics of sampled respondents are shown on Table 1.

Data analysis results
Analysis to evaluate the impact of technological innovation on a web-based electronic learning system was performed using a logistic regression approach. Correlation matrix was performed to find significant relationships among dependent variables. This could contribute to obtaining better results from the regression models. There is moderate to strong correlations, and the variables did contribute significantly (see Table 2) with no outlier(s).

The establishment of binary logistic regression model
The essence of using the logistic regression analysis method was to understand whether there is a linear or nonlinear relationship between dependent variables of innovative product and process of an electronic learning system (TI), usefulness of an electronic learning system (eLSuc), an outcome of an e-learning system (eLOC) and 42 independent variables (see Appendix A). The following represents the mathematical equation of a logistic regression: where ln {p/1-p} is the "log odds" of K, n represents the number of independent variables, p represents the proportion of successes, β 0 the constant (an intercept), β 1 (i = 1, …, n) the regression parameters (coefficient), X 1 to X 2 the independent variables (continuous or categorical) and e the error (or residual) of the equation.

Statistical tests of relationships between variables
The output of the binary logistic regression analysis of the dependent variables (TI, eLSuc, eLOC) and predictors that are statistically significant are given in Table 3. The output of statistical analysis between TI & eLSuc → eLOC and TI & eLOC → eLSuc are included in Table 3.
In Table 3, β is the coefficient of independent variable which is tested with Wald to identify the influence of each independent variable on dependent variable. The Wald measure on Table 3 shows TI predicts an event significantly since p < .05. The value of β = 4.012 signifies a change in Table 1 Including intercept in the model was important since the test was significant (p < .05). The following results are also identified on Table 3. The Odd ratio of 5.173 for TI → eLOC is slightly higher than that of eLSuc → eLOC (3.354) with p values (.000, .007) respectively, indicates each contributes significantly to eLOC. This implies, TI may contribute to eLOC 5 times compared to 3 times of eLSuc. PRI3 (p < .000, Odd ratio = 2.744) positively contributes the highest to TI; IQ6 (p < .001, Odd ratio = 4.377) positively contributes slightly higher than SeQ1, SQ3, and SQ4 to eLSuc; and BF6 (p < .000, Odd ratio = 3.025) positively contributes the highest to eLOC (see Appendix A).

Statistics of the overall model evaluation & goodness-of-fit (GoF)
The predictors consistently distinguished Yes from No of TI & eLOC → eLSuc (x 2 = 157.805, p < .000, df = 2) since a test of the constant only and full models was statistically significant. The Nagelkerke R 2 (54.1%) of the model explains the variance in eLSuc. For a GoF test, the .933 value of Hosmer-Lemeshow is insignificant (p > .05), indicating that the model is fit to the data. For TI & eLSuc → eLOC model, the chi-square = 64.562, p < .000, df = 2, and Nagelkerke R 2 .247 (24.7%) (see Table 4

Predictive accuracy and descriptive statistics
Classification tables were employed to predictive and evaluate the accuracy of the logistic regression model.

Predicting overall success
The overall success of 93.0% was predicted correctly (97.5% for "Yes" TI & eLOC impact eLSuc and 67.2% for "Not" impact eLSuc) as shown in Table 5. Thus, the model with predictors is improved and significantly better compared with the constant model only correct classification of 85.0%. On Table 5, 87.4% (95.3% for "Yes" TI & eLSuc impact eLOC and 41.3% for "Not" impact eLOC) was the overall success prediction. Thus, the model with predictors minimally and significantly predicts better than the constant model only correct classification of 85.3%.

Predicting a probability
Assignment rules of the indicator variable Y for the result of a test sample under the action of a set of independent variables are as follows: Y ¼ 1; TI or eLSuc or eLOC; 0; notTI or eLSuc or eLOC (2) where P is the probability of TI or eLSuc or eLOC occurrence and Q is the probability of nonoccurrence of TI or eLSuc or eLOC, the logistic regression's computational formula of P becomes: β 0 is the constant term unrelated to the factors x i , β 1 , β 2 , …, β m are regression coefficients which are the contributions of factor x i to P, and e (natural logarithms' base, approximately 2.72).
With formula P + Q = 1, we could get the formula to calculate the probability of non-occurrence of TI: Using the β values (logistic coefficients) in Table 3, the predictive calculation formula becomes: p ¼ e À 8:226þ4:012xTI ð Þ 1 þ e ðÀ 8:226þ4:012xTIÞ (5) Imagine a user who performs 3 e-learning activities per day within 1 h. Using the above predictive formula would determine whether the user decides that an e-learning system is useful (choosing YES). For a user to use an e-learning system to perform 3 important educational activities in 1 h would mainly depend on the innovative design, features, services, and methods of accessing that e-learning system. Substituting in the above assumptions, we get:  Hence, the likelihood that a user who completes three e-learning activities daily within an hour will decide an e-learning system is innovative is 97.8%. This affirms the correct prediction shown in Table 5.

Descriptive statistics
Using different e-learning systems over a long period of time has influence on determining whether an e-learning system is innovative or not. Hence, 52% ((225/380) * 87.9) of valid cases responded Yes to NoES (two, three), LoU (<1-2 yrs, >2 yrs), and TI (see Table 6) confirms that e-learning systems were innovative as respondents indicated (see Table 5).
Recommendation of an e-learning system to others reflects the intent of future use. The outcome (benefit) of using an electronic learning system determines its usefulness. 61% ((262/427) * 100) of valid cases responded Yes to eLSuc, eLOC, and RES (see Table 7), which confirms electronic learning systems were efficient and useful to the respondents (see Table 5).

Discussion and conclusion
The study focused on developing a context-based model to identify and evaluate specific factors of technological innovation that contributes to successful delivery, effective use, and positively impacts on learners of a web-based electronic learning system. Since a web-based electronic learning system is considered an information system, explicit and widely approved factors of technology acceptance and adoption, information systems success, and user satisfaction models were incorporated into the context-based model. Relationships and consequential effects among technological innovation (TI), an e-learning system (eLSuc) and the outcome of using an e-learning system (eLOC) were evaluated. Technological innovation impacts positively on the features of e-learning systems and the outcome of e-learning systems. Improvement in the features of e-learning systems results in positive outcomes of e-learning systems. The most important positive benefits of e-learning systems to students were risk minimization and achieving learning goals. NoES (Number of e-learning systems used), TI (Innovative products and processes), LoU (Length of using e-learning system) Improved software service is the factor that impacts positively and highest on web-based e-learning systems.
We used logistic regression method to analyze the relationships between TI, eLSuc, eLOC, and 42 independent variables. The Cronbach's alpha coefficient test result of .77 indicates reliability of the measurements. The variables did contribute significantly with moderate to strong correlations without outlier(s). Statistically, the model was fit to the data since the predictors consistently distinguished Yes from No of TI & eLOC → eLSuc (p < .000), the Nagelkerke R 2 explains 54.1% variance in eLSuc, and the Hosmer-Lemeshow test was p = .933. For TI & eLSuc → eLOC, the p < .000, Nagelkerke R 2 of .247, and the Hosmer-Lemeshow test of p = .889 suggest the model was appropriately derived from the data. The overall success of 93.0% (TI & eLOC impact on eLSuc) and 87.4% (TI & eLSuc impact on eLOC) were correctly predicted. H1 and H1a are statistically confirmed. However, TI impacts positively (p = 000, OR = 55.239, β = 4.012) on eLSuc higher than on eLOC (p = 000, OR = 5.173). This implies that, when TI level increases by 1 unit, features of eLSuc improve 55 times. This result is aligned with the findings of Alkandari (2015) and Wang et al. (2019) who are of the view that innovations have made e-learning systems more flexible, usable, interactive, and learning preferences. Technology innovativeness influences the acceptance of an electronic learning system (Campbell & Ma, 2015), and through an innovative technology interface, e-learning features such as system reliability, availability, and ease of use are improved (Al-Fraihat et al., 2020). For the positive impact of TI (β = 1.643) on eLOC, Damanpour (2010) asserts that pursuing technological innovation effort results in operational cost reduction and timely delivery of service that eventually maximize value for customers. Statistically, hypotheses H1b and H1c are confirmed. The negative β values indicate that decrease in the effort of product and process innovations would result in less innovative product and/or service. Improved software service contributed positively and highest (p = 000, OR = 2.744) to TI. Improved accessibility is not statistically confirmed (p = .495).
H2 is statistically confirmed (p = 007, OR = 3.354, β = 1.210). This implies that, when eLSuc level increases by 1 unit, eLOC improves 3 times. According to Tseng et al. (2015), ease of use and satisfaction of customers are improved when IT-facilitated innovative products and services are offered. H2a (SQ p = .032, IQ p = .009, SeQ p = .001) is statistically confirmed. According to Adel (2017), quality attributes impact the acceptance and adoption of an electronic learning technology which in turn improves the satisfaction of users and continues use of that technology (Dreheeb et al., 2016). The results also support findings of Cidral et al. (2018) where technical quality features of a system contribute to effective functioning, total satisfaction, and usefulness. Statistically, H2b (PU p = .003, PS p = .038) is confirmed. This result supports the findings of RES (Recommend e-learning system), eLSuc (Usefulness of e-learning system), eLOC (Outcome of e-learning system) Davis (1989) and Chen and Tseng (2012) that acceptance of an e-learning system is positively influenced by a perceived ease of use and usefulness. H2c (Use p = .040) is statistically confirmed. According to Davis (1989) and Van Raaij and Schepers (2008), the degree at which a learner accepts and use an electronic learning system determines the success of the system. Their findings support our result. An ease of use (p = 001, OR = 4.156), system understandability (p = .000, OR = 4.128), relevance of information (p = .000, OR = 4.377), and service responsiveness (p = .003, OR = 4.197) contributed the highest to e-learning success.
H3 is statistically confirmed (p = 007, OR = 3.354) where eLOC impacts positively on eLSuc. In education, online learning has yielded significant outcomes in relation to successful course delivery, effective use of systems, and positive benefits to learners (Al-Fraihat et al., 2020). Our findings affirm this. Statistically, hypothesis H3a is confirmed. Chang (2015) identified that e-learning leads to cost savings and improves learning while Yengin et al. (2011) found user satisfaction and net benefits as major determinants effective e-learning system. Our results corroborate with these findings. Achieving learning goals contributed highest (p = 000, OR = 3.025) to eLOC. Improving productivity is not statistically confirmed (p = .879).
Using different e-learning systems over a period of time results in achieving more benefits (Cidral et al., 2018) and has influenced on determining whether e-learning system is innovative and useful or not. 52% of valid cases responded of using two or three different forms of e-learning systems and for a period of over one year, hence e-learning systems used were innovative. 61% of valid cases responded that, usefulness and performance of e-learning system(s) are satisfactory, they have obtained important benefit(s) from using it, and would strongly recommend its use. This affirms the position of Kang et al (2018) who found that, usefulness, helpfulness, and overall satisfaction of an electronic learning system influence users to recommend usage.

Contributions of the study
This study aimed at identifying and evaluating specific factors of technological innovation that impact the usefulness of an e-learning system. Academically, this research impacts on the learning and teaching aspects of education and contributes to the ongoing research on technologymediated learning and teaching methods. The main components of IS success model by DeLone and McLean and Technology Acceptance Model 3 (TAM3) were incorporated into the developed and tested context-based model. This offers further empirical investigation of the model in a quest to obtain comprehensive understanding and advancing research on specific technology-mediated learning and teaching theories.
When conducting this research, the educational systems globally were affected by COVID-19, higher educational institutes had no choice than to use Online Learning Management Systems (LMS) (e.g., Sakai, Moodle, Blackboard), and there was high implementation cost associated with technology-mediated learning and teaching processes. The study provides insights on significant issues that could contribute to the improvement of the usefulness of electronic learning systems. For instance, technological innovations have made e-learning systems more flexible, usable, interactive and a learning preference as the study indicates. In addition, acceptance of online learning is influenced by technological innovations, and through an innovative technology interface, the accessibility, dependability, and ease of use features of an e-learning system are improved. Thus, there should be discrete initiative to exploit the unique capabilities of technological innovations to boost the functional characteristics of online learning systems and the ensuing benefits.

Research limitations and future work
Technological innovation was the focus of the study. We recommend including other categories of innovation and as well extend the investigation to other tertiary institutions in developed countries. Students were the participants for the study. Including other stakeholders (e.g., instructors, designers, implementers) in future research could enhance the findings. The proposed model could Appendix A.