In higher education, e-learning is increasingly being incorporated into face-to-face courses in the form of blended learning (Johnson, Adams, and Cummins 2012). Laurillard and Ljubojevic (2011) recommend that teachers beginning to teach using e-learning need to adjust their approach to learning and teaching, rather than simply transferring their approach from face-to-face teaching. Caudle and Moran (2012) highlight the importance of reflection when making this adjustment.

Goodyear (2005) has examined the significance of teachers’ philosophy, pedagogical beliefs, and teaching tactics and strategies in educational design. He recommends that when learning designs are developed by a team, it is important for team members to agree on a common set of beliefs and practices from the beginning, as differences in beliefs and practices can be problematic. Variance in teachers’ conceptions about how the web can be used by students to generate knowledge has been examined by González (2009). He identified a pedagogical hierarchy of three conceptions: use of the web for individual access to learning materials, information and individual assessment, use of the web for learning related communication, and use of the web for networked learning. Similarly, differences in the way teachers use blended learning have been identified by Lameras et al. (2012). They found teachers had different perspectives about the relationship between e-learning and face-to-face teaching in their units of study. Some teachers used the e-learning mode as a secondary environment for more teacher-centred teaching. Other teachers used e-learning as an equally important or primary environment for more student-centred teaching that emphasised the co-creation of knowledge and negotiated meaning.

e-Learning sustainability is defined by Stepanyan, Littlejohn, and Margaryan (2010) as ‘e-learning practice that evidently addresses current educational needs and accommodates continuous adaptation to change without outrunning its resource base or receding in effectiveness’ (p. 10). For the past decade, e-learning sustainability has predominantly been discussed from an institutional perspective. Early research focussed on the costs and benefits (JISCInfonet 2004), and scalability (Littlejohn 2003) of e-learning. Salmon, Jones, and Armellini (2008) presented an evidence-based model to improve scalability through a teamwork approach in which academic support staff, learning technologists and librarians work with university teachers to plan and develop courses using e-learning. In the field of Computer Supported Collaborative Learning, Pozzi and Persico (2013) have proposed the 4Ts model, designed to help teachers prepare and share learning designs, based on the elements of task, team(s), time and technology.

Robertson (2008) maintains that professional development addressing teachers’ e-learning pedagogy is crucial for encouraging change in teaching practices and the sustainable adoption of e-learning. Focusing on university teachers’ perspectives on the sustainability of ‘grass roots’ e-learning initiatives, Gunn (2010) argues successful e-learning initiatives need a learning design that has been proven to be beneficial, the potential to be adopted or adapted for use beyond the original context, and to be independent of the developers so ‘future prospects would not be compromised’ if they left (p. 90).

Although Gunn (2010) acknowledges the potential impact on e-learning sustainability when staff leave an institution, little research has examined this issue. An exception is research by Asensio, Whatley, and Jones (2001), which focused on a university teacher who took over a blended learning unit of study shortly before the semester started, after the previous teachers, who had developed the e-learning component, had left. Asensio, Whatley, and Jones highlight that e-learning designs are developed based on a teacher's beliefs; therefore, taking over another teacher's e-learning design means teaching with their beliefs, which may be dissonant with one's own and this may be problematic when the incoming teacher is not informed about those beliefs. It is this level of sustainability ‘at the chalkface’ that is the focus of this paper: what happens when teachers develop e-learning resources that are integrated into a unit of study that is then taken over by another teacher.

Although Hativa (2000) claims teaching practices need to change to improve teaching quality, after working with two university teachers rated as poor teachers by their students, she found other qualities also need modification if teaching is to improve: personal characteristics that impact on teaching (such as speech quality, insecure behaviour and an associative way of thinking), pedagogical knowledge, beliefs about teaching (such as teaching by knowledge transmission) and beliefs about students (such as attributing the full responsibility for learning to students). In describing the process of change in philosophies of teaching, Guskey (1986) found teachers change their practices, see the impact on learning, then change their beliefs. In contrast, Song and Looi (2012) found teachers change their beliefs before they change their practices. Taking another perspective, Caudle and Moran (2012) found fluidity in change in teachers’ beliefs and practices, with each one informing the other.

Looking at the impetus for change in teachers’ philosophies of teaching, Entwistle and Walker (2000) showed teachers change beliefs following development of their knowledge, experience and feelings, although only if existing beliefs are found to be inadequate. Similarly, Scott (2013) found teachers change their beliefs when their expectations are not met and they have the opportunity for collaboration and discussion with colleagues. Thus, change in beliefs may be stimulated by teachers’ unmet expectations or lesson images, described by Schoenfeld (1998) as teachers’ expectation of student work within a lesson. Mälkki and Lindblom-Ylänne (2012) examined the influence of reflection on change in teaching practices. They identified challenges that may constrain teachers, preventing them from engaging in reflection or changing their practice following reflection. These constraints may be organisational or curricula and include teaching with a curriculum that has been designed and prescribed by other teachers.

Although Stepanyan, Littlejohn, and Margaryan (2010) highlight the value of longitudinal research into e-learning sustainability, little longitudinal research has examined change in the e-learning beliefs and practices of university teachers taking over someone else's e-learning design. This study extends previous research by Asensio, Whatley, and Jones (2001), which focused on a technologically adept teacher who began teaching with an e-learning resource (taken here to mean online learning and teaching materials) developed by previous teachers. In that study, data was gathered over one semester through an interview with the teacher and access to her reflective diary. The focus of this study is a teacher who was inexperienced with e-learning and information technology, yet took over a blended unit of study with an e-learning resource developed by the previous teacher that accounted for one-fifth of the subject assessment and was directed towards academic skills development relevant to the degree program. What is distinctive about this study is the longitudinal approach. It examines the way the teacher changed the e-learning resource and its implementation over two years, and the way her reflections on the changes provided an opportunity and stimulus for change in e-learning beliefs and practices.

This is one of six case studies from a multiple case study focusing on university teachers in health and associated disciplines. Yin (2009) recommends case study design when ‘a ‘how’ or ‘why’ question is being asked about a contemporary set of events over which the investigator has little or no control’ (p. 13). The qualitative approach taken enabled the researcher to focus on the people involved and incorporate her knowledge of the context and her insights and perceptions, with the goal of drawing upon and extending existing theory (Neumannn 2003). A longitudinal design was used in data collection to identify and comprehend development over time (Burns 2000). This research falls within the constructivist framework, using a subjectivist epistemology, where the researcher and participant co-create understanding using naturalistic methodologies (Denzin and Lincoln 2011). The research questions are: 1) What changes does the teacher make to the e-learning resource and the way it is used? 2) How does the teacher's reflections about the e-learning resource influence the changes she makes? 3) How do the teacher's experiences, reflections, practices and beliefs influence one another?

The case study was conducted over two non-concurrent semesters at a research-intensive, campus-based Australian university (hereafter ‘the University’). Using purposive sampling, a participant was selected who implemented an e-learning resource for the first time in 2007 and re-implemented it in 2008.

Semi-structured interviews of 40–50 minutes were held with the participant at the beginning and end of the semesters in which the e-learning resource was used. Each was recorded and transcribed and field notes were taken during the interviews. At the beginning and end of each implementation, the e-learning resource was analysed to identify what students were expected to do with it, the level of demand and assistance provided, and aims and learning objectives. Unit of study documents were analysed and a 20-minute semi-structured interview was held with the departmental information technology support officer. Thus, the participant's descriptions of her e-learning resource and student use were substantiated through other forms of data collection, facilitating trustworthiness and crystallisation of data (Tracy 2010).

A case study database was used to collect data and form a chain of evidence to identify conclusions and increase trustworthiness (Yin 2009). Preliminary data analysis during the data collection phase tested and extended existing theory. This focused on the categories and broader concepts identified in the research literature, which formed the basis of the research instruments (Bazeley 2009). As new categories and concepts were identified, the research instruments were revised. Final analysis occurred at the end of data collection.

Approval was granted by the University's Higher Research Ethics Committee. In this article, the names of the participant, her colleagues, department and institution have been changed. Excerpts from interviews include the interview number, i.e. interview 1 as (i1).

Three challenges for Jennifer concerned her inexperience with e-learning and information technology, the iterative process of e-learning development and the learning design of the e-learning resource. Over two years, Jennifer's reflections on these challenges led to elaborations of her e-learning beliefs and practices.

Due to the limited handover in taking over the previous teacher's e-learning design, Jennifer experienced many difficulties in the first implementation of the e-learning resource. This formed the basis of her reflections on her experiences and changes in her beliefs and practices.

Jennifer reflected that her inexperience with e-learning and lack of confidence using e-learning technology challenged her adoption of the e-learning resource. Consequently, she was reluctant to make major changes before the first implementation. Jennifer's hesitancy is in line with research by Mahdizadeh, Biemans, and Mulder (2008), which found teachers’ attitudes to computers and the web influence the way they use them in their teaching. Similarly, in studies of pre-service teachers using social networking software in teaching, Turvey (2012) found those who use information technology extensively outside their studies are more confident using it in their teaching.

Although Jennifer's e-learning resource had been developed based on another teacher's teaching goals and e-learning beliefs and practices, these were only communicated to her in the way they were embodied in the resource. At the beginning of the study, Jennifer was inexperienced with e-learning, so her own e-learning beliefs and practices were mostly based on her experience with face-to-face teaching. Therefore, in line with research by Caudle and Moran (2012), she based her revision of the e-learning resource before the first implementation on her face-to-face experience. This became problematic when, part-way through the first implementation, she realised she needed to make changes to the e-learning resource and these changes were not communicated to students and affected assessment items, causing confusion.

Following the first implementation and student evaluation, Jennifer redeveloped the learning design of the e-learning resource and her teaching goals. She elaborated her e-learning beliefs and practices, particularly those concerning the relevance of activities. However, following significant changes before the second implementation, Jennifer found she had not provided sufficient guidelines to students. She realised students need clear communication and guidelines while working on e-learning resources alone and off-site. At the end of the second implementation, Jennifer also noted that she needed to provide specific criteria to students to promote work at a higher cognitive level. Thus, the iterative nature of e-learning development, identified by Phillips, Kennedy, and McNaught (2012), posed additional challenges.

The finding that Jennifer changed her e-learning beliefs before her e-learning practices confirms research by Song and Looi (2012), who found change in beliefs is required before change in practices. In this study, Jennifer changed her beliefs, then changed her practices, observed the results, then confirmed her beliefs. This also supports research by Entwistle and Walker (2000), who found teachers’ beliefs can change through development of knowledge, experience and feelings if existing beliefs are found to be inadequate.

The changes in Jennifer's beliefs were influenced by ‘critical unmet expectations’, a term introduced by Mälkki and Lindblom-Ylänne (2012). Jennifer's students’ use of the e-learning resource did not match her expectations, or lesson image (Schoenfeld 1998), of important learning outcomes. Scott (2013) found teachers are more likely to change their beliefs following critical unmet expectations if they have the opportunity for discussion and collaboration with colleagues. Yet, while Jennifer could discuss and collaborate with her current co-teacher, she could not do so with the teacher who had developed the e-learning resource, who could also be considered a co-teacher.

After the second implementation, Jennifer discussed her plans to introduce online discussions for her students to communicate while undertaking their practical component. This indicates a change in e-learning beliefs and practices from providing structured information and students individually studying materials provided by the lecturer, towards those in which ‘the Web is used for learning related communication’, identified by González (2009).

Jennifer did not understand the relevance of some components of the e-learning resource because she did not have discussions with the previous teacher about her e-learning beliefs and practices, which underpinned the learning design of the e-learning resource. Before the first implementation, Jennifer said the previous teacher ‘may have had something in her mind she forgot to tell’ her. Stein, Shephard, and Harris (2011) have highlighted that a teacher's e-learning beliefs inform their e-learning practices. Thus, by taking over another teacher's e-learning resource, a teacher is taking over their e-learning beliefs and practices.

One other study, by Asensio, Whatley, and Jones (2001), reported a similar finding when a university teacher took over a unit of study with an e-learning resource, with no opportunity for discussion with the previous teachers. Like Jennifer, the teacher had little time to review the e-learning resource before implementation. Asensio, Whatley, and Jones highlight the difficulties when taking over another teacher's e-learning resource because ‘Taking over someone else's design implies taking over someone else's representation of knowledge … and their personal understanding of the students’ intended learning experience’ (p. 73).

Because of the lack of handover and time for review, Asensio, Whatley, and Jones (2001) found the new teacher had to take the content and pedagogical design of the e-learning resource ‘at face value’ and hoped to make changes once the semester started (p. 69). However, during the semester, the teacher only had time to try to understand the design of the e-learning resource and how assignments were marked, and deal with pressing technical and managerial issues. Similarly, Jennifer had little time to change the e-learning resource substantially during the first implementation.

The difficulties involved in taking over someone else's e-learning design have been substantiated in studies by Rambo (2012) and Archambault and Crippen (2009), who examined school teachers using pre-packaged courseware incorporating e-learning resources. The teachers in the studies reported difficulties when there were errors in the course materials they were unable to correct. In contrast, Archembault and Crippen noted that the teachers reported a positive experience when they could create their own content.

When investigating scalability of e-learning resources, Littlejohn (2003) found an e-learning resource is more likely to be used by the teacher who developed it, rather than other teachers. This has implications for the success of the current Open Educational Resources (OERs) movement. Similarly, Gunn (2010) contends that university teachers have difficulty using other teachers’ e-learning resources because of a ‘not invented here’ syndrome. Taking a different perspective, Goodyear (2005) highlights the challenges that may occur when teachers working together have different beliefs and practices. He claims that when teachers who work in teams developing e-learning resources have different beliefs, it ‘can lead to fatal divergence’ because beliefs directly influence practices (p. 86). Viewed from Goodyear's perspective, Jennifer and the previous teacher were co-teachers with divergent e-learning beliefs and practices.

Because Jennifer's e-learning beliefs differed from the previous teacher's, she felt constrained by the framework of the e-learning resource and was critical of it, substantiating research by Mälkki and Lindblom-Ylänne (2012) that teachers lose motivation when there are elements in their teaching context they cannot change. After redeveloping the e-learning resource, Jennifer was more positive about it because, according to Mälkki and Lindblom-Ylänne, she had been ‘able to apply [her] own understanding and values, as well as see the results of [her] own actions’ (p. 42). Nevertheless, Jennifer was still constrained by the goals of the e-learning resource because it formed part of a series throughout the course.

Asensio, Whatley, and Jones (2001) highlight that it is more difficult to take over an online course than a face-to-face course because an online course is usually ‘set in stone’, available to students from the start of semester (p. 69). Because of this, e-learning resources must be finalised before the semester begins, which can be challenging for teachers taking over a unit of study with an e-learning resource just before the semester starts. Mahdizadeh, Biemans, and Mulder (2008) emphasise the time needed for e-learning, in which teachers need to learn how to use the technology, then prepare to use it in their teaching.

In discussing the poor uptake of e-learning by university teachers, Littlejohn (2003) argues that university teachers need support for the complex nature of designing and developing e-learning resources. She recommends this should include technical systems, as well as organisational arrangements and people to provide design and technical support. This study highlights that similar support and time is needed for teachers taking over another teacher's e-learning resource so they can redevelop it according to their e-learning beliefs and practices. Time and support is also required when e-learning resources created by others are implemented.

Gunn (2010) recommends a team-teaching approach to e-learning design to ensure continuity following staff changes. Jennifer's e-learning resource had been developed by the previous teacher working with a team on a series throughout the course. While they provided an overview of the series, it would have helped Jennifer if a teacher other than the one who left had been involved in developing the e-learning resource. This teacher could have informed Jennifer about the underlying beliefs. The team-based approach to e-learning development recommended by Salmon, Jones, and Armellini (2008) also emphasises the importance of staff discussion and documentation of underpinning beliefs and goals. Similarly, the 4T Model developed by Pozzi and Persico (2013) to assist in the design of Computer Supported Collaborative Learning by focussing on task, teams, time and technology facilitates the sharing of designs. Finally, Adams (2004) recommends teachers record their beliefs when designing e-learning resources to assist other teachers using them. This highlights the need for a well-structured handover of e-learning resources, including documentation, which could include notes about the underlying philosophy of teaching and learning, the rationale for the development of the e-learning resource and the learning objectives. It would be useful if teachers could easily add to the documentation each time they use the resource.

It would have been useful to capture Jennifer's change in beliefs over the two years of data collection through a self-report log, which would have picked up other influences on her beliefs and practices between implementations of the e-learning resource in non-concurrent semesters. However, self-report logs are time-consuming for participants and, given the two year timeframe, it proved difficult to recruit when this method of data collection was proposed to prospective participants. This is an example of the compromises needed to conduct meaningful research in naturalistic settings, which may have been more difficult given the context of a research-intensive university.

Although the research reported here provides insight into the challenges in taking over another teacher's e-learning design, it is based on one case study. Additional research could aim to generalise these findings across a larger number of university teachers, disciplines, teaching modes and institutions.