Predicting Academic Staffs Behaviour Intention and Actual Use of Blended Learning in Higher Education: Model Development and Validation

This sub-section reviews prior studies that have employed UTUAT and TPACK models to examine factors that influence students and lecturer use of BL and e-learning as shown in Tables 6 and 7 in the appendix respectively. Tables 6 and 7 further presents the aim, methodology applied, and study context. Evidence from the reviewed 28 studies suggest that UTAUT and TPACK have been successfully employed. However, none of the reviewed studies have integrated both UTAUT and TPACK to investigate BL. Thus, this study adopts both UTAUT and TPACK, where UTAUT is utilized to examine the factors that influence lecturers’ behavioral intention and actual BL usage. Furthermore, the socio-demographic factors that may determine lecturers’ behavioral intention and actual use of BL are examined.

The Malaysian government like other countries is advocating the use of technology to facilitate educational activities in higher educations (Anthony et al., 2019). Thus, the Malaysian government has initiated a national council for lifelong learning committee to delivers the platform for coordination e-learning facilities and formulating policies to support e-learning (Chea et al., 2012). Similarly, the Malaysian Educational Blueprint (2015–2025) founded by the Ministry of Higher Education (MOHE) Malaysia termed Globalized Online Learning (GOL) to promote innovative and creative use of Information Communications Technology (ICT) in teaching and learning strategies in public or private institutions (Anthony et al., 2019). MOHE encourages universities, colleges, and polytechnics to implement digital methods of teaching and learning course materials to enhance the skills and knowledge of students based on ICT mediated instructional design for teaching and learning (Chea et al., 2012).

As such, many lecturers in Malaysia higher education are encountered with challenges to design novel interactive syllabus and also deploy BL approaches to facilitate their teaching pedagogy to resolve the needs of learners thus allowing students to have better learning with the course content (Bokolo Jr et al., 2020). Likewise, lecturers are also using virtual learning tools to access online learning resources thus allowing learners to become self-learners and more proactive (Sivapalan, 2017). Moreover, BL helps to facilitate their teaching styles while keeping learners engaged and motivated with the content which helps create active and distributed learning environments (Al-shami et al., 2019). Currently, based on MOHE Malaysia institutions are to implement BL based on the formula, “Blended course = (T1 >  = 1) + (TR >  = 7) + (TA >  = 3) = (TAS >  = 2)” to assess their BL effectiveness. Where TI is course information, TR is course resources, TA is course activities and TAS is course assessment. Thus, MOHE Malaysia set a target, where 40 per cent of total course offered should be blended (Anthony et al., 2019).

A number of theories have been adopted to investigate lecturers’ intention and acceptance to use new technologies in educational context such as Technology Acceptance Model (TAM), Unified Theory of Acceptance and Use of Technology (UTAUT), Diffusion of Innovation (DoI) Theory, etc. (Scherer et al., 2019). UTAUT model was developed by Venkatesh et al. (2003) to explore and empirically compare components from different technology acceptance models in predicting and explaining use behavior of variables that influence technology adoption behavior over time (Saleem et al., 2016).

UTAUT was proposed based on eight technology acceptance theories which comprises of social cognitive theory, motivational model, model of personal computer utilization, TAM, DoI, theory of reasoned action (TRA), Theory of Planned Behavior (TPB), and combined TAM and TPB (Venkatesh et al., 2003). The UTAUT model consists of six constructs (performance expectancy, effort expectancy, social influence, facilitating conditions, behavioral intention and use behavior) (Venkatesh et al., 2003). Moreover, UTAUT model comprise of four moderating (gender, age, experience, and voluntariness of use) as seen in Fig. 1.

Figure 1 depicts the UTAUT model. The main constructs are described below;

UTAUT model has been employed in e-learning domain by prior studies (Abu-Al-Aish & Love, 2013; Al-shami et al., 2019; Dečman, 2015; Kocaleva et al., 2014; Lakhal et al., 2013; Lwoga & Komba, 2015; Saleem et al., 2016; Sattari et al., 2017; Tarhini et al., 2017) to examined factors that influence students’ acceptance of e-learning. Likewise, UTAUT model has been employed by previous BL studies (Brand et al., 2011; Gawande, 2015, 2016; Khechine et al., 2014; Radovan & Kristl, 2017) that investigated students and/or lecturers’ intention to accept BL. In this study UTAUT was selected as one of the models to examine factors that influence lecturers’ behavioral intention to use BL because UTAUT includes a social construct which is important in BL environment as mentioned by Dečman (2015).

Based on the study termed “Those Who Understand: Knowledge Growth in Teaching” Lee Shulman (1986) proposed the model of Pedagogical Content Knowledge (PCK). Shulman (1986) highlighted that there is need for a more comprehensible theoretical model regarding what lecturers should know and be competent to do, asking essential questions which relates to the categories and domains of content knowledge in the minds of lecturers and how general pedagogical knowledge and content knowledge inter-relates (Archambault & Crippen, 2009; Shulman, 1987). Thus, based on PCK, the technological pedagogical content knowledge (TPACK) model was developed by Mishra and Koehler (2006), to support lecturers towards effectively integrating technology in teaching based on technological pedagogical and content knowledge referred to as TPACK (Alsofyani et al., 2011).

TPACK aims to provide understand and describe the kinds of knowledge required by lecturers for effective ICT deployment for teaching (Alsofyani et al., 2012). In BL context TPACK highlights the significance of preparing academic staffs to make practical choices in their utilization of technology when teaching specific course content to a specific group of students (Scherer et al., 2019). As it can result to a better understanding regarding how lecturers make decisions that influence BL integration and acceptance into teaching and learning activities (Qasem & Viswanathappa, 2016). TPACK model comprises of content knowledge (CK), pedagogical knowledge (PK), technological knowledge (TK), pedagogical content knowledge (PCK), technological content knowledge (TCK), technological pedagogical knowledge (TPK), and technological pedagogical content knowledge (TPACK) (Alsofyani et al., 2012) as seen in Fig. 2.

Figure 2 depicts the TPACK model, thus each of the seven components are described below;

Furthermore, findings from prior BL studies (Alsofyani et al., 2012; Papanikolaou et al., 2017; Yang & Chen, 2010) suggested a positive outcome in adopting TPACK to improve lecturers’ capability to use ICT for teaching practice. Likewise, finding from the literature (Anderson et al., 2013; Maor, 2017; Maor & Roberts, 2011; Ward & Benson, 2010) suggested that TPACK guide teachers’ effort in addressing issues related to teaching and learning that results from the rapid changing technologies. Hence, TPACK model is adopted in this study to investigate the factors to be employed by lecturers' that influence actual BL usage in higher education.

This study considers the gender, age, ICT experience of the lecturers and voluntariness of use of BL by the lecturers as socio-demographic factors or moderating variables. Note that a moderating variable is a quantitative or a qualitative variable that influences the strength and/or direction of the relationship between two other independent variables (Anthony et al., 2020b; Baron & Kenny, 1986). Venkatesh et al. (2003) stated that gender, age, experience, and voluntariness of use are factors that moderate the relationship between the independent variables (performance expectancy, effort expectancy, social influence, and facilitating conditions) and behavioral intentions to use variable.

Findings from the literature (Kocaleva et al., 2014; Lakhal et al., 2013) also provide strong evidence for the significant effect of the moderating variables effects. Accordingly, Abu-Al-Aish and Love (2013) reported that gender and age moderated students’ behavioral intentions to use mobile learning. Also, findings from Lakhal et al. (2013); Khechine et al. (2014) indicated that the gender and age of students mediates behavioral intentions to use BL, stating that females were found to be more profound than males and therefore the impact on behavioral intentions was higher for females, predominantly for older females.

Likewise, findings from Venkatesh and Morris (2000); Padilla-Meléndez et al. (2013); Dečman (2015) indicated that both men and female technology acceptance decisions were influenced by their use behavior. However, regarding age, men use of technology diminished as their age increases. Similarly, results from Abu-Al-Aish and Love (2013); Khechine et al. (2014) suggested that in BL adoption male students are more concerned with the learning performance whereas female students are more interested with the ease of use and social influence factors opinions regarding the use of BL. Also, empirical evidence from Venkatesh et al. (2003) suggested that the impact of performance expectancy on behavioral intentions was higher for younger people, but the impact of social influence and effort expectancy were much lower for older people. Furthermore, regarding age, gender and computer experience influence on BL adoption, findings from Gawande (2015) indicated that older students (adult students) need more assistance and technical support than the young, aged students in relation to using BL approached for learning, known as facilitating conditions.

This is in line with results from Brand et al. (2011) where the authors mentioned that the young age students achieved higher learning performance in using iPad for BL approaches as compared to older students. Findings from previous studies (Abu-Al-Aish & Love, 2013; Brand et al., 2011; Dečman, 2015; Gawande, 2015; Kocaleva et al., 2014; Saleem et al., 2016) revealed that lecturer perception towards BL have a significant impact on their prior experience on technology usage. Lastly, based on the original UTAUT model (Venkatesh et al., 2003) predicted the influence of voluntariness of use on user behavior intention. Besides, the results of the research of Kocaleva et al. (2014); Saleem et al., (2016), conducted in e-learning environment, support the influence of voluntariness of use on student/teacher behavior intention to use e-learning systems. Therefore, similar to prior research (Vankatesh et al., 2003; Abu-Al-Aish & Love, 2013; Kocaleva et al., 2014; Saleem et al., 2016) the effect of gender, age, ICT experience, and voluntariness of use are examined in this study. Accordingly, we suggest the following hypotheses that;

Based on the factors that influence lecturers’ behavior intention and actual use of BL grounded by UTAUT and TPACK model, and socio-demographic factors, the proposed model is developed as seen in Fig. 3.

Figure 3 depicts the proposed model based on UTAUT and TPACK model to examine the factors that impacts lecturers’ behavioral intention and actual BL usage in higher education.

This study adopts a quantitative research method and data was collected from academic staffs from Malaysia institutions (see Table 1) that adopts both F2F weekly classes and e-learning system such as Massive Open Online Courses (MOOC), Learning Management Systems (LMS), etc. in teaching and learning process. Moreover, each lecturer involves in the data collection process uses e-learning system as a platform to upload teaching material (e.g., links, power point,.doc,.pdf files, etc.), publish course schedule/timetable, carryout course description and summary, manage class assignments submitted, conduct online quiz, and conduct class management in their respective institution.

Links to the survey was sent to purposive sampled selected academic staffs in Malaysia institutions, where the target sample for this research included academic staffs who teach blended courses. The lecturers were selected due to their experience and familiarity with blended settings in providing answers based on their perceptions toward BL environment. Accordingly, implicit consent was provided to the respondents who completed the survey. The aim and purpose of the study and respondent’s rights not to partake in the survey was clearly specified. Hence, participation in the survey was voluntary.

The pre-validated questionnaire in English language and was sent to experts (7 IT and 3 education domain) to proofread and amend the questionnaire for face and content validity to verify the correctness of the questionnaires in the pre-test phase, after which pilot study was carried out and data was collected from 10 lecturers to assess if the respondents understand the questions and to test the reliability of questionnaires instruments. Accordingly, Table 8 in Appendix depicts the variables, items, and sources. The questionnaire comprised of 51 items, and at the end of the data collection a total of 544 samples was collected, but 131 samples were excluded due to incomplete data which resulted to a final 413 usable samples. Table 9 in Appendix depicts the demographic characteristics of the survey respondents. Then, invitations to participate in the main survey, including weblink to the questionnaire, was distributed to lecturers via emailed and the weblinks was also distributed to the faculties and e-learning centers of the selected universities, colleges, and polytechnics from January 2019 to March 2019. The data collection involves a survey of academic staffs in Malaysia institutions as seen in Table 1.

On average, each respondent took not more than 10 min to answer all questions. The questionnaire included demographic question (gender, age, nationality, job title, qualification, institution type, institution category, years of experience in teaching, years of involvement in e-learning, years of experience in ICT, employment type, educational field, and voluntary number of bl training attended 2016–2018) measured using ordinal measurement (see Table 9).

Although some rewards were given to some participants based on a raffle draw for respondents that provided their emails address after completion of the survey. All research ethics were adhered to when coding, analyzing, and reporting of the collected data. All response from the survey remained anonymous to avoid any form of bias.

This study employed Statistical Package for Social Science (SPSS) version 23 and Structural Equation Modeling (SEM) based on Partial Least Squares (PLS) for data analysis. SEM approach was selected because it can be utilized to analyze all hypotheses in a single analysis (Lin & Wang, 2012). Similarly, SEM was employed in this research due to its ability to analyze the relationships between the variables and approximate random errors in the observed constructs directly in providing precise measurements of the questionnaire items and variables (Teo, 2019). Furthermore, PLS is a latent SEM technique that uses a component-based method for estimation (Anthony Jr, 2019). Thus, PLS-SEM provides two analyses which include assessment of measurement model (evaluation of reliability and validity of constructs) and assessment of structural model (checks relationships among model variables) (Hair et al., 2016). SmartPLS version 3.0 was deployed to assess the measurement and the structural model and SPSS version 23 was employed to carry out descriptive analysis.

This is the first step involved in assessing the developed model (see Fig. 3). This step helps to assess how well the observed questionnaire items measure the unobserved variables as presented in Table 8 (Teo, 2019). The measurement model was evaluated based on descriptive analysis, item loadings, reliability measures, convergent validity, and discriminant validity.

This is the second step which involves the test of the structural model which confirms the relationships among the variables (Teo, 2019). Accordingly, the model hypotheses (H1–H12) (see Fig. 3) are tested by deploying PLS algorithm in SmartPLS 3.0 based on bootstrap re-sampling performed to examine the path significance levels of each hypothesis. Results from Table 5 depicts the hypotheses testing, where statistical significance of each hypothesis was assessed based on a two-tail test (***). Additionally, the structural model assessment is measured by examining the path coefficients value (β) which evaluates the association between variables based on their degree of significant levels (p value) which is significant when p =  < 0.05. Moreover, the coefficient of determination termed R² value is used to measure the predictive significance of the model hypotheses. Next, the t-value is employed to assess the effects of each hypothesis which is based on the regression coefficients and associated significances as listed in Table 5 and Fig. 4, where t-value should be greater than 1.96 (Hair et al., 2016).

Results from Table 5 and Fig. 4 show the hypotheses test using a two-tailed t-test with a significance level of 5% (0.05). As seen all t-values are higher than 1.96. Furthermore, results from Table 5 also depict the β and R² values which is the different path coefficients ranking of the hypotheses, where H1 has the strongest effect of 0.905 (0.820), followed by H5 with 0.889 (0.791), then H8 with 0.760 (0.578), next is H4 with 0.846 (0.715), then H12 with 0.746 (0.557), H11 with 0.738 (0.544), H10 with 0.733 (0.538), H3 with 0.733 (0.537), H2 with 0.719 (0.517), H6 with 0.717 (0.515), H9 with 0.708 (0.501) and lastly H7 with 0.685 (0.469). Therefore, the hypothesized path relationship (H1-H12) is statistically significant since the β and R² values are greater than 0.1 and p-values are lower than 0.05 (Hair et al., 2016).

To validate the effect of gender, age, experience, and voluntariness of use on lecturers’ behavioral intention to use BL approach for teaching, we employ multi-group analysis (MGA) in SmartPLS3 as suggested by Hair et al. (2016); Padilla-Meléndez et al. (2013), to confirm the hypotheses H13a, H13b, H13c, and H13d. MGA was previously adopted by Anthony et al. (2020b). Thus, this study opted to validate if they exist a structural relationship between behavioral intention and use behavior in relation to the moderating variables to ascertain if there is a change in the variance values of gender, age, experience, and voluntariness of use. Thus, each moderating variable is tested twice. Results from MGA are shown in Fig. 5, where the first test depicts the path coefficients value (β) and the second test shows the t-value which should be higher than 1.96 to be significant.

Results of the moderating effect path coefficient of gender on lecturer’s behavioral intention to use BL is given as β = 0.010 which should be lower than “0” (Anthony et al., 2020b) and t-value is given as 0.262 which is lesser than 1.96 (see Fig. 5). The results suggest that, the gender do not mediate lecturers’ intention to use BL. Thus, reject H13a, since no significant was observed. In addition, the moderating effect path coefficient of age on the lecturers is given as β = 0.021 and t-value is given as 0.995 also lower than 1.96. Thus, indicating that age does not mediate lecturers’ behavioral intention to use BL. Hence, reject H13b. Similarly, the moderating effect path coefficient of ICT experience on the lecturers is given as β = -− 0.036 (see Fig. 5), whereas the t-value is given as 1.325 which is lower than 1.96. Thus, suggesting that prior ICT experience of lecturers does not mediates their intention to use BL, hence reject H13c. Likewise, the moderating effect path coefficient of voluntariness of BL is given as β = 0.055 (see Fig. 6), whereas the t-value is given as 1.593 also lower than 1.96. Hence, indicating that the lecturers’ voluntariness of using or not using BL to teach does not have any effect on their behavior and intention to use, thus reject H13d.

Accordingly, researchers such as Chin (1998); Anthony et al. (2020b) mentioned that researchers should not only confirm if there is a significant relationship among the variables or not, but also check the size of effect between the constructs. Thus, in addition to confirming hypotheses there is need to check the model’s effect size which measures the percentage of the importance and impact levels of the identified factors influence on the use behavior intention of lecturers to use BL approaches for teaching \({R}^{2}\) value for both UTAUT and TPACK model. Where, the effect size assesses the strength of correlation among the UTAUT and TPACK constructs. This helps in providing answer to the last research question, this confirming the complete impact of the study. Hence, importance-performance map test (IPMA) was deployed similar to prior BL study (Bervell & Umar, 2018) as seen in Fig. 7 to test for total effect for importance (values inside the circles) and impact levels (values on the line) of the identified factors.

Hence results from IPMA test in SmartPLS3 as depict in Fig. 7 reveal the test of total effect for UTAUT suggesting that performance expectancy is the most influencing driver with total effect of 0.905 in relation to behavior intention of lecturers towards BL. This result confirm findings from prior studies (Lakhal et al., 2013; Venkatesh et al., 2003) which stated that performance expectancy is the most influential UTAUT construct. Additionally, results from Fig. 7 show that effort expectancy is the most performing driver with value of 75.346, signifying that the perceived academic performance to be derived from adopting BL approaches will impact lecturers’ attitude towards BL. Considering TPACK, pedagogical knowledge has the higher effect of 0.221 and is also the most performing determinant with value of 72.532. Hence, institutions should lay more emphasis on the pedagogy employed for BL in teaching students as lecturers’ behavior to use BL is influenced by their pedagogical knowledge in managing blended course contents, teaching, and learning resources.

This study develops a model to examine the factors that influence lecturers’ behavioral intention and actual BL usage based on the applicability of UTAUT and TPACK model in Malaysia higher education. Moreover, this study investigated how gender, age, experience, and voluntariness of use moderate lecturers’ intention and actual use of BL. This research has successfully supported both theoretically and empirically applicability of UTAUT and TPACK as useful theoretical models for better understanding lecturer’s behavioral intention to accept and use BL for teaching in Malaysia higher education. Data was collected using survey from lecturers in universities, colleges, and polytechnics and analyzed using PLS-SEM. Grounded on UTAUT model. The results suggest that performance expectancy positively influences lecturers’ behavior intention towards using BL for teaching. This result is consistent with findings from prior studies (Abu-Al-Aish & Love, 2013; Lakhal et al., 2013), where the authors stated that performance expectancy relates to extent to which academic staffs assumed that their teaching performance would influences their perception towards using BL for teaching activities. Likewise, findings from Radovana and Kristl (2017) revealed that the performance expectancy impact of BL predicts if lecturers will find BL useful when they teach at their convenience and quickly towards improving teaching productivity.

The results also indicated that effort expectancy positively influences lecturers’ behavior intention towards using BL for teaching. This driver is considered as an important construct of teaching behavioral intention to use BL approach (Tarhini et al., 2017). This result confirms prior works (Saleem et al., 2016; Sattari et al., 2017), in which the findings from the authors suggested that effort expectancy which refers to the extent to which lecturer believes that BL adoption will have self-efficacy and will be easy to use in implementing teaching activities. Furthermore, findings from prior study Lwoga and Komba (2015) demonstrated that effort expectancy is a significant factor that affect both involuntary and voluntary environment during the initial stages of BL adoption. Findings from this study suggest that statistical significance was observed between social influence and lecturers’ behavior intention towards using BL for teaching. This suggests that lecturers would agree to continue using BL due to both internal and external influence such as from the university management and government pressure (Dečman, 2015; Sattari et al., 2017). Besides, other studies (Brand et al., 2011; Gawande, 2015; Khechine et al., 2014) also reported similar finding stating that academic staffs are influenced to adopt BL for educational purpose if it is essential that others believe he/she should use BL approaches.

The study also found that facilitating conditions significantly predicts lecturers’ behavior intention towards using BL for teaching. This result suggest that lecturers will continue to use BL approach if infrastructure are provided to support BL adoption such as the provision of training, facilities, and support to lecturers (Radovana and Kristl, 2017). The result of this study is consistent with findings from previous studies (Abu-Al-Aish & Love, 2013; Sattari et al., 2017) which indicated that the quality of service of BL provided to lecturers can influence their behavioral intention and level of acceptance towards the acceptance of BL. Thus, it is evident that the better the support perceived by the lecturers, the more they will continue to use BL (Kocaleva et al., 2014). Further, this study found that behavioral intention represents lecturers' intention of using BL approaches for educational purpose, which is consistent with results reported by Saleem et al. (2016); Sattari et al. (2017) which revealed that behavioral intention relates to lecturers' intention of using BL approaches in the future for teaching and learning activities.

Based on TPACK model, statistical results indicate support that technology knowledge of lecturers has a significant influence on their behavior to use BL for teaching. This result suggests that TK entail lecturers understanding on how to utilize various technologies for teaching (Graham et al., 2009). This result is analogous with findings from the literature (Koehler & Mishra, 2009; Schmidt et al., 2009) which specifies that lecturers’ knowledge of operating systems, software, and hardware, and the capability to utilize teaching software applications such as Microsoft word, PowerPoint, Excel spreadsheets, creating of documents, use of browsers, and e-mail will impact their use of BL for teaching. Furthermore, the results suggest that content knowledge of lecturers predicts their behavior towards using BL for teaching. According to Koehler and Mishra (2009) CK refers to the knowledge lecturers must be aware of the content they intent to teach and how the content of that knowledge differs for several content areas. This implies that CK relates to fundamental theories, concepts, facts, and procedures of the actual subject matter that is to be taught or learned (Sahin, 2011). Hence, findings from prior studies (Alsofyani et al., 2011; Qasem & Viswanathappa, 2016) revealed that lecturers’ knowledge of how to organize and connect course content ideas for BL determines if they will accept BL.

Pedagogy knowledge is also found to have a strong effect on lecturers’ behavior towards using BL which confirms the studies carried out by (Alsofyani et al., 2012; Papanikolaou et al., 2017; Yang & Chen, 2010). Respectively, PK involves lecturers’ knowledge about the practices or procedures of teaching and learning and how it links to educational aims and objectives. Moreover, lecturers’ PK involves issues related to student learning, managing classroom, developing and implementing lesson plan. This result also corroborates findings postulated in the literature (Wang, 2004; Schmidt et al., 2009), that PK entails knowledge steps to be followed by the lecturers in the classroom based on the type of students and strategies for assessing student learning thus influences teacher’s acceptance of BL. In addition, the results indicate that pedagogy content and knowledge predict lecturers’ behavior towards using BL. This result is similar with findings from other BL studies (Alayyar et al., 2012; Maor & Roberts, 2011), which mentioned that PCK entails the intersection of pedagogy and content. Therefore, PCK represents the integration of pedagogy and content by the lecturer into an understanding of how the subject matter are prearranged, adapted, and represented for teaching students in a BL environment (Koehler & Mishra, 2009; Schmidt et al., 2009).

On the other hand, technology content knowledge appeared to have a significant influence on lecturers’ behavior towards using BL. This result is in line with finding from prior studies conducted by Tømte et al. (2015), where the authors suggested that TCK relates to lecturers’ perception and understanding of how adopting a particular technology can change the way students learn and practice in a particular content area. Furthermore, results from other studies (Graham et al., 2013; Koehler & Mishra, 2009) confirmed the result from this study that TCK which is knowledge relating to the method in which content and technology for teaching entails how technology is used by lecturers to present course contents to students. The results also revealed that technology pedagogy knowledge impact lecturers’ behavior towards using BL for teaching. This result seems quite reasonable and consistent with findings of previous studies (Wang, 2004). In order words, TPK refers to lecturers’ knowledge of how several technologies can be deployed in teaching and understanding that utilizing such technologies may change the way the lecturer teaches (Lye, 2013). TPK is lecturers’ knowledge of the current modules, and capabilities of different technologies that can be deployed for BL teaching (Koehler & Mishra, 2009). Thus, it involves lecturers’ knowing how teaching with a particular technology might change the learning outcome of students (Ward & Benson, 2010). Additionally, the results indicate that technological pedagogical content knowledge of lecturers positively impact on their behavior to use BL for teaching. This is evident because TPACK encompasses the knowledge lecturers need for incorporating technology into their teaching (Wang, 2004). Thus, results from Schmidt et al. (2009) revealed that lecturers must have a spontaneous understanding of the composite interplay between CK, PK, TK components of knowledge by teaching course content using suitable pedagogical strategies and technologies (Koehler & Mishra, 2009).

Furthermore, regarding the moderating effects of gender, age, experience, and voluntariness of use on lecturers’ intention and actual use of BL, the results reject the effect of gender moderating lecturers’ intention to use BL for teaching. This result is in line with results from Saleem et al. (2016), where the authors found that gender do not influence the acceptance of Moodle as a teaching and learning tool. But is dissimilar to the findings from Lakhal et al. (2013); Padilla-Meléndez et al. (2013); Dečman (2015); Gawande (2015), who revealed that gender influences e-learning adoption. Conversely, the results also reject that the age of lecturers moderates their intention to use BL for teaching. This contradicts findings from previous studies (Venkatesh & Morris, 2000; Venkatesh et al., 2003; Lakhal et al., 2013; Khechine et al., 2014; Saleem et al., 2016). Regarding influence of experience, the results reveal that prior IT experience of the lecturers does not moderate their intention to use BL for teaching. This result is analogous with results from the literature (Dečman, 2015) who stated that experience does not have a positive influence student intention to use e-learning. Lastly, voluntariness of use is proven not to moderate lecturers’ intention to use BL for teaching. On the contrary, this result is not consistent with findings from prior study (Saleem et al., 2016).

This study provides academic staffs in higher education with guidance for how to implement BL approach based on empirical data regarding factors that influence lecturers’ behavioral intention to use BL derived from UTAUT model (Venkatesh et al., 2003) and factors to be employed by lecturers' that influence actual BL usage derived from TPACK model (Koehler & Mishra, 2009). Therefore, this research is a one of the first studies that integrates UTAUT and TPACK to develop a model to examine technology acceptance literature in BL environment. Overall, the results of the present study suggest that UTAUT and TPACK models were able to provide empirical evidence of lecturers’ acceptance and use of BL approach for teaching. Theoretically, based on UTAUT model this research provides valuable insights into technology acceptance and use in academic setting. Overall, this research adds to the body of knowledge about acceptance theory in BL environment. Thus, the utilizing of UTAUT model as a base theory in the context of lecturers’ behavioral intention to adopt BL for teaching provide possible prospects for future research on the investigation of various theoretical perspectives to understand lecturers’ BL continued usage intentions. The model developed in this study provide better understanding into how institutions can develop initiatives to encourage continuous usage of BL strategies for lifelong learning and future studies among lecturers towards planning and implementing a successful BL approach in their respective institutions.

Moreover, findings from this study are useful to e-learning administration to identify important factors to be considered in developing appropriate strategies and policies to promote long-term usage of BL approaches in higher education. Additionally, the main outcomes of this study are to develop a model that better understand the factors that affect academic staffs’ behavioral intention to use BL by academic staffs in institutions. This study concludes that performance expectancy, effort expectancy, social influence, and facilitating conditions play important roles on lecturer’s behavioral intention to use BL. Another important contribution of this study is that socio-demographic factors (gender, age, experience, and voluntariness of use) as moderating variables do not influence lecturers’ behavioral intention to use BL for teaching. The results of our study confirm that lecturers’ behavior intention is an important variable that predicts their attitude to use BL.

Respectively, findings from this study help to better understand the perception of lecturers in Malaysia towards BL adoption which can help educators, policy makers, and practitioners to understand what the lecturers expect from BL approaches. This can help institutions administration in achieving the most effective deployment of BL and also aid them to improve their strategic decision making regarding technology for education in future. Moreover, this study provides the current practice of BL in Malaysia higher education from the lens of lecturers’ opinions on the important factors that impact the acceptance and adoption of BL approaches which will help in improve future direction of BL practices. Furthermore, the institutions should formulate policies to guide BL usage and development by focusing on educating lecturers to use BL through training, workshops, printed leaflets or posters, and electronic medium such as social media, institutions’ website, etc. For instance, it should be mandatory for lecturers to deliver course modules, assignment, assessment and feedback to learners by using both F2F and online medium. Thus, BL policy should stress the prominence of collaboration between faculties and department in providing BL infrastructure, IT technical help desk and support to students and lecturers.

The findings of this study have practical implications for academic staffs and course designers in proving a model that serves as a basis that support lecturers in integration ICT in improving teaching and learning performance. Ultimately, TPACK is incorporated in the developed model to serves as a valuable tool for assessing lecturers’ knowledge in the domain of technology integration to improve teaching efficiency. Besides, as there are fewer empirical research on the applicability of TPACK in improving BL in Malaysia educational context, the current study will be useful for educators, academicians, and lecturers to provide a roadmap to improve BL experiences towards developing alternative and new approaches as compared to the conventional methods of teaching. Furthermore, based on TPACK, the developed model provides an analytical instrument for assessing what lecturers should knowledge regarding BL adoption. Specifically, findings from this study helps to incorporate the use of technology in providing important implications for examining issues related to BL use. The instrument developed as presented in Table 8 can be employed by academic staffs in universities, colleges, and polytechnics as a benchmark tool to improve the acceptance and use of BL approaches to improve teaching and learning effectiveness and also assist lecturers in focusing on important aspects of effective teaching in both F2F and online mode.

Findings from this study offer several possibilities for promoting research in academic staffs’ education, lecturer professional development, and lecturers’ use of technology by developing a model that examines the factors to be employed by lecturers' that influence actual BL usage. Thus, this study provides a better understanding of lecturers’ views of knowledge in relation to course content, pedagogy and technology use in improving teaching. The model can significantly be used by academic staffs to monitor and improve their current BL activities in measuring their knowledge about teaching regarding teaching improvement and career development. As a reflective tool, the model provides an in-depth analysis method for lecturers to improve teaching in BL environments. Moreover, university administration can use the develop model based on TPACK items as seen in Table 8 to assess lecturer practices and perceptions towards BL. Thus, the developed model serves as a useful instrument in analyzing actual on-campus and F2F teaching practices of lecturers for the continuous motivation and development of academic staff teaching in Malaysia higher education. By employing TPACK, the developed model assists the development of BL approaches for discovering and describing how technology-related knowledge is instantiated and implemented in practice.

This research has some limitations. First, the result from this study is from Malaysia higher educations, thus generalizability of the results to other countries should be treated with caution. Secondly, this study employs quantitative survey to collect the data. While, the survey is theoretical valid and reliable, adopting only quantitative approach limits the ability to have an in-depth investigation which is mostly found in qualitative method. Thirdly, this study was conducted among lecturers in management and engineering faculties where the use of BL approaches is widespread. Fourth, the effectiveness of BL in teaching was explored and student learning outcome was not examined in this current study. Therefore, future studies could replicate this study within other countries. Moreover, qualitative case study by interview can be adopted to further explore the model to understand BL acceptance and adoption. Also, academic staffs in more disciplines should be included for comparative analyses. Additionally, individual variables such as teaching and teaching satisfaction, etc. should be included to enhance the predictive significance in explaining the impact of BL for teaching and learning effectiveness. Lastly, students learning outcome will be examined in future studies.