Investigating university students’ intention to use mobile learning management systems in Sweden

ABSTRACT

This study aims to explore university students’ intention to use mobile learning management systems (m-LMS) in higher education in Sweden. To address this, a research model based on the technology acceptance model (TAM) and nine research hypotheses are developed. In addition to perceived usefulness and perceived ease of use from the TAM, the research model includes academic relevance, university management support and perceived mobility value from previous literature as external variables to study the adoption of m-LMS. The proposed research model and research hypotheses are empirically tested with 130 university students in Sweden. According to the results, support is found for seven of the nine research hypotheses. This study contributes to the validation of the extended TAM for mobile learning and demonstrates the hypothesised model moderately predicts students’ intention to adopt m-LMS in higher education in Sweden.

KEYWORDS:

• Mobile learning management system (m-LMS)
• mobile learning
• technology acceptance model (TAM)
• academic relevance
• university management support

Introduction

Mobile learning is a fast-growing trend in educational settings as the development of mobile technologies has enabled learning on the move (Han & Shin, 2016). Sharma and Kitchens (2004) defined mobile learning as new type of learning supported by mobile devices including ubiquitous communication technology and intelligent user interfaces. The emergence of mobile learning allows students to enjoy personalised learning on their mobile devices. In the past few years, we have seen many new mobile services which integrate mobile technologies into universities’ educational systems (e.g., (Gao, Krogstie, & Siau, 2014)). Han and Shin (2016) found that the use of mobile learning was associated with higher exam scores.

As mobile learning evolves, the emergence of various applications including the existing learning management system (LMS) – could be accessed from mobile devices, which is called a mobile learning management system (m-LMS) (Brantes Ferreira, Zanela Klein, Freitas, & Schlemmer, 2013). The m-LMS is a kind of mobile learning tool that provides faculty and student access to courses on their mobile devices regardless of time and location.

Although mobile learning management systems provide a means for university students to assess learning information with their mobile devices, there is a lack of studies on students’ intention to use these systems. Providing an m-MLS does not guarantee that students would actually use it in their daily lives. Students might have different perceptions of the system. For instance, some students were not aware of the potential value of using the system before they started using it (e.g., (Nguyen, Barton, & Nguyen, 2015)). To the best of our knowledge, there is no empirical study on university students’ adoption of m-LMS in Sweden. Therefore, it is worth exploring this increasing trend. Sweden is one of leading countries in the world for good Internet infrastructure and high mobile phone penetration rate. University students in Sweden can easily access many mobile services with a good wireless network infrastructure.

The purpose of this research is to examine students’ intention to use an m-LMS (i.e., Mobile Blackboard) in higher education in Sweden. To address this, a research model based upon existing technology acceptance theories is developed and nine research hypotheses are accordingly proposed. This research model is empirically tested using survey data collected from 130 university students in Sweden.

Literature review

Technology adoption theories

Several models have been developed to test the users’ attitude and intention to adopt new technologies or information systems. These models include the Technology Acceptance Model (TAM) (Davis, 1989), Theory of Planned Behaviour (TPB) (Ajzen, 1991), Innovation Diffusion Theory (IDT) (Rogers, 1995), and Unified Theory of Acceptance and Use of Technology (UTAUT) (Venkatesh, Morris, Davis, & Davis, 2003). Among the different models that have been proposed, TAM, which is the extension of the Theory of Reasoned Action (TRA) (Fishbein & Ajzen, 1975), became the key model in understanding the predictors of human behaviour toward potential acceptance of the technology (Marangunić & Granić, 2015). TAM focuses on the perceived usefulness (PU) and perceived ease of use (PEOU) of a system and has been tested in some domains of E-business and has proven to be largely reliable at predicting user acceptance of some new information technologies such as electronic commerce (Pavlou, 2003), and online shopping (Gefen, 2003).

Technology adoption in educational settings

Although TAM was originally conceived as a model to explain technology adoption by users in business and commercial environments, it has been further investigated as an appropriate research model set in an educational context (Alharbi & Drew, 2014; Hsu & Chang, 2013; Sánchez & Hueros, 2010). For instance, Sánchez and Hueros (2010) extended the TAM to include technical support and perceived self-efficacy to study students’ adoption of Moodle. The results indicated that technical support has a direct effect on perceived ease of use and usefulness.

Among the various previous studies that have applied TAM or extended TAM in technology adoption, many have tended to focus on the adoption of mobile learning (e.g., (Huang, Lin, & Chuang, 2007; Park, Nam, & Cha, 2012) and LMS (e.g., (Alharbi & Drew, 2014)). These studies provide us with insights into the factors that might influence the adoption of m-LMS. Alharbi and Drew (2014) studied the adoption of LMS with an extended TAM and found job relevance positively influenced perceived usefulness and perceived ease of use. Huang et al. (2007) employed an extended TAM to explain the acceptance of mobile learning. The results demonstrated the importance of perceived enjoyment and perceived mobility value (PMV) to an individual’s acceptance of mobile learning. Iqbal and Bhatti (2017) also found university support as an external variable of m-learning adoption. The existing findings above provide a foundation to develop the research model and hypotheses presented in Section 3.

Adoption of m-LMS

Although many previous studies applied TAM or extended TAM to examine the adoption of e-learning and mobile learning, as presented in last section, little research has been conducted to examine the factors influencing the adoption of m-LMS. Some existing studies on the adoption of m-LMS are summarised as follows. Han and Shin (2016) examined both factors influencing the use of m-LMS by online students and the effect of using m-LMS on online students’ academic achievement. The findings indicated age and employment status were significant factors in predicting students’ adoption of m-LMS. Akman and Turhan (2017) examined the users’ acceptance of social learning systems in higher education with an extended TAM. They found that perceived usefulness, security awareness and ethical awareness were significant predictors of users’ behaviour towards using social learning systems. Joo, Kim, and Kim (2016) investigated factors predicting online university students’ actual usage of an m-LMS in Korea and found that PU and satisfaction predicted users’ continued intention to use m-LMS, but PEOU was not related to it.

Research model and hypotheses

Research model

Based on the facts supporting the TAM and its success in including additional variables to better understand factors related to adoption in educational technologies, we develop an extended TAM model (see Figure 1) to study the adoption of m-LMS. In this study, PU is defined as the degree to which undergraduate students believe that using an m-LMS would enhance their academic conduct. In the regard, PEOU is defined as the judgment with respect to the amount of effort required in using an m-LMS. Behavioural Intention of use (BI) indicates the strength of inclination of the students towards using m-LMS.

Figure 1. Research model.

Beside TAM’s core constructs defined above, three key moderators, Perceived Mobility Value (PMV), Academic Relevance (AR) and University Management Support (UMS) are derived from previous studies to investigate students’ intention to use m-LMS within a university education context. The concept of perceived mobility consists of three distinct dimensions of human interaction: spatial, temporal and contextual mobility (Kakihara & Sørensen, 2001). Since mobile technology has become integral to the increasingly mobile nature of people’s lifestyle, mobility is accordingly perceived as a critical advantage of mobile learning. PMV allows users access to services/information anywhere at any time via mobile devices (Huang et al., 2007). Thus, PMV in this study denotes the understanding and consciousness of university students regarding the existence of PMV as an advantage for using m-LMS.

The selection of AR in this research is based on the core theoretical argument underlying the role of cognitive instrumental processes of Job Relevance (Venkatesh & Davis, 2000) and Compatibility (Rogers, 1995) namely that users form PU judgments in part by cognitively comparing to what extent a new technology is capable of working according to their expectations. AR is similar to the concept of job relevance proposed in extended TAM (Venkatesh & Davis, 2000). Job relevance is defined as the degree to which an individual believes that the target system is applicable to his or her job (Venkatesh & Davis, 2000). The setting of this research is to study students’ adoption of m-LMS in their academic lives. Thus, we proposed the use of AR for this research. We define AR as being that students form PU judgments by reasoning how relevant or compatible an m-LMS would be for their academic lives.

Lastly, university management is responsible for delivering a good user experience for its users. UMS would therefore be a significant antecedent to examine students’ intention to use an m-LMS. This study defines UMS as an institutional commitment and interest that ensures technology needs to be up to date, but also sufficiently mature or stable for its success and continuation (McGill, Klobas, & Renzi, 2014).

Research hypotheses

We have developed nine research hypotheses based on the research model. Each hypothesis, as labelled in Figure 1, is briefly described below.

Hypotheses concerning TAM variables

The TAM is proposed following the relationship between its constructs: BI is positively affected by attitude towards (ATT) usage and PU, ATT is positively affected by PU and PEOU, and PU is directly affected by PEOU. These relationships have been proven in previous studies on mobile learning (e.g., (Huang et al., 2007; Iqbal & Bhatti, 2017; Park et al., 2012)). Thus, we propose the following hypotheses.

H1: PEOU positively affects PU of m-LMS.

H2: PEOU positively affects attitudes towards using an m-LMS.

H3: PU positively affects BI to use m-LMS.

H4: PU positively affects attitudes towards using an m-LMS.

H5: ATT positively affects BI to use m-LMS.

Hypotheses concerning external variables

Perceived mobile value (PMV)

Previous research has investigated the positive effect of PMV on PU (Huang et al., 2007) for mobile learning. We believe that effect of PMV would also be present in users’ perceived usefulness of m-LMS. Based on this, we posit the following hypothesis.

H6: PMV has a significant impact on PU of m-LMS

Academic relevance (AR)

In this study, AR means relevance of m-LMS in university education in general. Venter, van Rensburg, and Davis (2012) investigated the determinants of usage of an online learning management system in South Africa and found that Major Relevance had a positive influence on PU. Moreover, Venkatesh and Davis (2000) found that Job Relevance had a positive direct effect on PU. Consequently, this study argues AR, in general, affects the PU of m-LMS. Therefore, this study proposes the following hypothesis.

H7: AR positively affects the PU of m-LMS.

University management support (UMS)

In the previous research, Iqbal and Bhatti (2017) signified the importance of UMS for mobile learning initiatives. They also found that UMS positively influenced students’ PEOU and PU of mobile learning. Furthermore, Venkatesh (2000) also concluded that facilitating conditions positively influenced PEOU and PU. Barker, Krull, and Mallinson (2005) also indicated that the university support staff played an important role in the day-to-day support and maintenance of the mobile learning infrastructure within their learning institution. Thus, we propose the following hypotheses.

H8: UMS positively impacts on the PU of m-LMS.

H9: UMS positively impacts on the PEOU of m-LMS.

An empirical study

In this empirical test, the proposed research model and research hypotheses are examined for the adoption of Mobile Blackboard with 130 university students in Sweden. Blackboard (a.k.a., the Blackboard Learning Management System), is a virtual learning environment and course management system developed by Blackboard Inc. Mobile Blackboard is Blackboard’s mobile solution that helps students stay informed, up to date and connected with their daily studying activities. Therefore, Mobile Blackboard can be regarded as a kind of mobile learning management system.

Survey instrument

The survey instrument consists of two parts. The first part incorporates nominal scale to identify respondents’ demographic characteristic information, while the second is intended to examine users’ adoption of m-LMS. To ensure validity, the questionnaire used in this study has been modified from the original measurement scales used in TAM (Davis, 1989) and from other relevant literature (Huang et al., 2007; Iqbal & Bhatti, 2017; Park et al., 2012; Venkatesh & Davis, 2000). In order to ensure that the instrument better fits the context of m-LMS, some minor word changes were made to ensure easy interpretation and comprehension of the questions. The questionnaire has been written in English. Moreover, to avoid problems that may occur in wording, measurement and ambiguities, the questionnaire was pre-tested by a native English speaker.

As a result, 16 measurement items¹ are included in the second part of the survey. A 7-point Likert scale, with 1 being the lowest score (strongly disagree) and 7 being the highest (strongly agree), was used to examine participants’ responses to all items in this part. In addition, data are analysed using the structural equation modelling.

Sample

The survey was distributed to undergraduate students in the school of business in one of the universities in Sweden in the form of paper-based questionnaires in May 2017. The participants were briefed about the purpose of survey and provided information on m-LMS before participation. On average, the participants took 4 minutes to complete the survey. As a result, a total of 190 students voluntarily participated in the survey, of which 130 responses were valid. The participants have also been informed that the results would be reported only in aggregate and that their anonymity would be assured. Of the participants, 49% were male, and 51% were female.

Measurement model

The quality of the measurement model is determined by (1) Content validity, (2) Construct reliability and (3) Discriminant validity (Bagozzi, 1979). To ensure the content validity of our constructs, a pre-test with one researcher in information systems was carried out. We found that the questionnaire was well understood by the researcher.

To further test the reliability and validity of each construct in the research model, the Internal Consistency of Reliability (ICR) of each construct was tested with Cronbach’s Alpha coefficient. As a result, the Cronbach’s Alpha values range from 0.829 to 0.907. A score of 0.7 is marked as an acceptable reliability coefficient for Cronbach’s Alpha (Robinson, Shaver, & Wrightsman, 1991). All the constructs in the research model were equal to or above 0.70. Consequently, the scales were deemed acceptable to continue.

Convergent validity was assessed through Composite Reliability (CR) and the Average Variance Extracted (AVE). Bagozzi and Yi (2012) proposed the following three measurement criteria: factor loadings for all items should exceed 0.5, the CR should exceed 0.7, and the AVE of each construct should exceed 0.5. As shown in Table 1, all constructs were within acceptable ranges.

Table 1. Factor loadings, composite reliability, and AVE for each construct.

Construct	Item	Factor Loading	CR	Cronbach’s Alpha	AVE
PEOU	PEOU1	0.886	0.90	0.84	0.76
	PEOU2	0.909
	PEOU3	0.813
PU	PU1	0.848	0.89	0.82	0.73
	PU2	0.800
	PU3	0.904
ATT	ATT1	0.883	0.87	0.70	0.77
	ATT2	0.869
BI	BI1	0.927	0.92	0.83	0.86
	BI2	0.925
AR	AR1	0.878	0.89	0.74	0.80
	AR2	0.904
UMS	UMS1	0.947	0.94	0.88	0.90
	UMS2	0.944
PMV	PMV1	0.898	0.91	0.81	0.84
	PMV2	0.933

The measurements of discriminant validity are presented in Table 2. According to the results, the variances extracted by the constructs were more than the squared correlations among variables. This fact revealed that constructs were empirically distinct. As good results for convergent validity and discriminant validity were achieved, the test result of the measurement model was good.

Table 2. Discriminant validity.

	AR	ATT	BI	PMV	PU	PEOU	UMS
AR	0.891
ATT	0.620	0.876
BI	0.719	0.706	0.926
PMV	0.479	0.731	0.637	0.916
PU	0.641	0.746	0.650	0.626	0.851
PEOU	0.467	0.606	0.550	0.592	0.544	0.870
UMS	0.592	0.678	0.691	0.695	0.586	0.475	0.946

Note: Diagonals represent the average variance extracted, while the other matrix entries represent the squared correlations.

Structural model and hypotheses testing

The structural model was tested using SmartPLS. Figure 2 presents the structural measurement model. Table 3 presents the path coefficients, which are standardised regression coefficients. Seven (H2, H3, H4, H5, H6, H7, H9) of the nine research hypotheses were significantly supported.

Table 3. Test of hypotheses based on path coefficient.

Hypothesis	Path coefficients	Result
H1: PEOU->PU	0.150	Not Supported
H2: PEOU->ATT	0.285**	Supported
H3: PU-> BI	0.276**	Supported
H4: PU->ATT	0.588**	Supported
H5: ATT-> BI	0.501**	Supported
H6: PMV->PU	0.294**	Supported
H7: AR-> PU	0.382**	Supported
H8: UMS->PU	0.082	Not supported
H9: UMS->PEOU	0.475**	Supported

*P < 0.05 **P < 0.01.

Figure 2. Structural measurement model.

The R² (R square) in Figure 2 denotes coefficient of determination. It provides a measure of how well future outcomes are likely to be predicted by the model, and the amount of variability of a given construct. According to the results, the amount of variance in the user intention to use m-LMS explained by the model was 0.53. The explained variance of perceived usefulness factor is 56%. The percentages of variance explained for PEOU and attitude to use m-LMS are 23% and 61% respectively.

Discussion

According to the results, PU had a significant positive influence on BI and attitude toward use. These findings are consistent with the previous studies on technology adoption (Venkatesh & Davis, 2000) and in the context of mobile services (Gao et al., 2014). However, PU showed a relatively weak direct relationship with BI compared to its strong influence on ATT. This may be due to the fact that students have more reliance on the LMS version already accessible to them on their laptops, hence perceptions of use are not entirely real but rather influenced.

The results indicated that PU, in turn, was positively influenced by PMV and AR. This is consistent with results from previous research (e.g., (Huang et al., 2007; Venkatesh & Davis, 2000)). This is also consistent with the findings from previous studies (e.g., (Mallat, Rossi, Tuunainen, & Oorni, 2006)) that mobility was closely related to PU in terms of increased performance offered by the technology. UMS, in this study, showed no significant impact on PU, which is inconsistent with the findings from previous studies (e.g., (Barker et al., 2005)). Both PMV and AR together explained moderate variance of (56%) in PU. It predicted that 56% of the students’ perception of usefulness of mobile learning management system would be derived from the m-LMS’s ubiquitous ability to deliver learning experience and especially due to the m-LMS’s importance and relevance during their studies.

PEOU had shown no influence on PU, which is inconsistent with the TAM’s original relationship and other previous studies on technology adoption (e.g., (Chang, Yan, & Tseng, 2012)). A plausible reason might be that the students in this study did not perceive ease of use as a critical factor that would improve their academic performance in the process of using m-LMS, which is consistent with one of the findings in (Hsu & Chang, 2013). Another possible reason might relate to the subjects in this study. Undergraduate students in Sweden were relatively advanced in using  and experiencing mobile technology. Furthermore, UMS had a strong relationship with PEOU and it moderately explained variance of (23%) in PEOU. PEOU, in turn, showed a direct impact on ATT.

The structural model also indicated that PEOU and PU have a significant relationship with ATT, and both constructs together explained the 61% variance in the attitude toward use, which further strongly influences behavioural intention to use m-LMS. ATT predicted 56% variance in BI, showing a moderate intention to use m-LMS during studies in higher education.

From a theoretical perspective, this research is considered to be beneficial for academic research since it extends and enhances the understanding of the adoption of mobile learning. Specifically, this research contributed to the current literature on the adoption of m-LMS. This study expanded the boundary conditions of the technology acceptance model into the emerging mobile learning context and demonstrates the robustness and explanatory power of the hypothesised TAM framework to predict users’ intention to adopt an m-LMS in a higher education context. Consistent with prior research, PU and PEOU were the major determinants of users’ attitudes toward using the m-LMS app, which, in turn, is of primary importance to their intention to use it. External factors such as AR and UMS also had strong positive impacts on PU and PEOU of using m-LMS.

The implications of this study went beyond the validation of the theoretical model. The findings helped us understand how university undergraduate students will interact and value the use of an m-LMS and offer insight into designing and implementing mobile technologies for educational purposes. Students confirmed that academic relevance and perceived mobility value in m-LMS were significant factors for its usefulness and hence m-learning experience. University management support is critical to ensure competent technical people who will, in turn, ensure a design of m-LMS that is easy to use. Difficulty in navigating downloading, searching and sharing from a small screen of a handheld device may affect students’ interest to use an m-LMS. Similarly, the findings would be useful for the developers and service providers of m-learning applications such as m-LMS in order to understand the key factors that would play a critical role in their successful adoption in higher education.

However, we were also aware of some limitations of this work. Firstly, we only tested the research model and hypotheses with university students in Sweden. Therefore, the generalisability of the results to other countries remains to be determined. Secondly, this study may be limited due to the relatively small sample size used. Thirdly, the research model constructed for this study included only three external variables within the scope of the technology acceptance model and failed to study other contextual information, such as users’ prior experience and types of mobile learning activities. Last but not least, all the data are collected using self-reported scales. This may lead to some caution because common method variance may account for some of the results.

Conclusion and future research

This study was designed to explore the university students’ behavioural intention to use m-LMS in a higher education context. A research model based on TAM and nine research hypotheses were presented, and empirically tested with a sample of 130 university students in Sweden. In addition to PU and PEOU from TAM, the model includes AR, UMS and PMV from previous literature as external variables to study the adoption of m-LMS. Analysis revealed significant relationships in seven out of nine hypotheses.

This study contributed to the validation of an extended TAM model for a mobile learning context and demonstrates the hypothesised model moderately predicts students’ intention to adopt an m-LMS in a higher education context. Relationships between the core constructs of TAM with the exception of PEOU on PU were consistent with prior research. Similarly, UMS on perceived usefulness relationship was also not significant. However, UMS showed a strong relationship with PEOU and PEOU, in turn, revealed direct influence attitude toward use. By comparison, PU remained the dominant mediating factor to derive students’ intention to use m-LMS through ATT. The results also revealed usability problems in the app. For instance, some students indicated that it was difficult to navigate and search information on Mobile Blackboard due to non-compatibility of the interface of the m-LMS on mobile devices. To the best of our knowledge, this is among the first studies on m-LMS adoption; notably, there is no prior study in Sweden on this subject. Therefore, the findings could help researchers, practitioners and many online education service providers to use this study as a starting point.

Continuing with this stream of research, we plan to examine the adoption of m-LMS from the teachers’ perspective. Future research is also needed to empirically verify the research model with larger samples in other universities of users’ adoption of m-LMS. A longitudinal study (i.e., a 3-month period) is another opportunity to re-validate the research model. By choosing a longitudinal method, the research can more closely examine the change in user behaviour over time. Such a study can help to enhance the understanding of how and why students use m-LMS over time. Last but not least, future studies may extend the model with other external variables relevant to examine the adoption of m-LMS in higher education.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Asher Irfan Saroia

Mr. Asher Irfan Saroia holds a MSc in information systems from department of informatics, school of business, Örebro University, Sweden. His research interests include mobile learning, the adoption of information technology, usability evaluation, the use of learning platforms in higher education, personal health record, human computer interaction, and User-Human centered design.

Shang Gao

Dr. Shang Gao is an assistant professor at department of informatics, school of business, Örebro University, Sweden. His research interests include in mobile information systems, digital business, digital ecosystem, enterprise modelling, blockchain technology, and technology diffusion. He has published more than 60 refereed papers in journals, books and archival proceedings since 2006.

Notes

1. The measurement items are available at: https://tinyurl.com/y8ymxgum.