Understanding fully online teaching in vocational education

Literature has previously reported that student-centred practices are the mark of good pedagogy in online education. In contrast, the competency-based nature of vocational education in Australia has been understood to encourage teacher-centred pedagogy. The likely tensions between these two teaching contexts are not yet understood, and little is yet known about the pedagogy of fully online vocational education teachers. To begin understanding pedagogy in this context, a wide-ranging digital survey was implemented. Findings revealed that online vocation education teachers conceived good online pedagogy as student-centred, yet student-student learning opportunities were rated lower than teacher-student practices. Notably, enacted practice was consistently more teacher-centred than teachers’ ideal, and factors within the teaching context were perceived by teachers as a limitation. They reported their workload to be dominated by marking and administration ahead of student-centred practices such as building rapport. This work is of interest to researchers and institutions navigating the continued expansion of online education and the ongoing demand for effective student-centred practice.


Introduction
It has been reported that teachers generally teach how they themselves were taught (Ellis & Hafner, 2003), and until recently, most of their education experience will have been in face-to-face classroom contexts (Brinthaupt, Fisher, Gardner, Raffo, & Woodard, 2011). However, online education is growing significantly across sectors (Sun & Chen, 2016), and there are reports that teachers are challenged when their familiar face-to-face teaching strategies are not able to be effectively enacted online (Banas & Velez-Solic, 2013). In essence, experienced teachers once again become novices during the shift to online education (Redmond, 2011), and online education is a teaching context that is understood to challenge existing conceptions about the teaching role (Kilgour, Reynaud, Northcote, McLoughlin, & Gosselin, 2019). As teachers navigate from classroom to online teaching contexts, they reconceptualise how to connect their teacher self with the different affordances and limitations of teaching online (Baran, 2011). This includes a reconceptualisation of their beliefs about what represents 'good' pedagogy in the different teaching context that is online education (Rodrigues, Almeida, Figueiredo, & Lopes, 2019). To understand pedagogy in any teaching context, Jensen, Price, and Roxå (2019) have recommended that it is important to understand the interplay between teachers, their practice, and their teaching context. This study investigates those three parts within an under-researched online teaching context-vocational education (Chang, 2016;Griffin & Mihelic, 2019). It contributes new knowledge about the nature of online vocational education and factors that result in enacted practice in this context being more teachercentred than teachers conceived as ideal.
To develop a background for the presentation of this study, the relationships between conceptions of teaching, teaching context, and enacted practice are described next, including a brief summary about teacher-centred and student-centred classifications of pedagogy. Then, this study is introduced and situated within a wholly online vocational education teaching context. The survey instrument design and analysis is then described ahead of survey results being presented and discussed in three parts: (1) teacher characteristics and their relationship with enacted practice; (2) teaching profiles and relationships, alignments, and misalignments between conceptions of teaching and enacted practice; and (3) teaching context factors that affect enacted practice. The conclusion then summarises key outcomes, and directions for ongoing research are proposed.

Background
In this section, key concepts relevant to this study are introduced. These include relationships between conceptions of teaching, teaching context, and enacted practice, and the classification of pedagogy as teacher-centred or student-centred.
Teacher's beliefs about pedagogy are known as their conception of teaching. Conceptions of teaching represent the way a teacher is inclined to teach, or would prefer to teach, within a particular teaching context (Clark & Peterson, 1986;Samuelowicz, 1999). This is important because conceptions of teaching have long been known to directly affect enacted practice (Conrad, 2012) and the relationships between these two parts of pedagogy have been an important element of education research for three decades (Harshman & Stains, 2017). Enacted practice is different from conceptions of teaching in that enacted practice describes the teaching actions and interactions actually undertaken by the teacher (Gibbons, Villafañe, Stains, Murphy, & Raker, 2018;Westbrook et al., 2013). Importantly, enacted practice has been found to not always align with conceptions of teaching (Doyle, Seery, Canty, & Buckley, 2019). This is because enacted practice is also affected by the perceived affordances or limitations of each teaching context (Clark & Peterson, 1986). This means that the real or perceived affordances, limitations, or needs of a particular teaching context can cause teachers to shift their practice toward or away from their ideal (Eley, 2006;Samuelowicz & Bain, 1992). For example, Shi, Delahunty, and Gao (2018) found that large class sizes, an exam-driven teaching context, as well as the teacher's own confidence, experience, and knowledge, hindered the enactment of some idealised pedagogies. So important are the relationships between conceptions of teaching, teaching context, and enacted practice, that Kennedy (2010) recommended no understanding of pedagogy in any context can therefore be developed without considering all three parts.
Notably, both conceptions of teaching and enacted practice can be classified as teacher-centred or student-centred (Eley, 2006). Since the seminal work of Samuelowicz and Bain (1992), the nature of teacher-centred and student-centred pedagogy has been much researched and debated for both face-to-face and online teaching contexts. Presenting the deeper review of those classifications is beyond the scope of this study. However, in summary, within face-to-face classrooms, teacher-centred pedagogy has been described as teachers transmitting information, which is often curriculum-bound, to students in structured ways and this may include handouts and giving examples from the teacher's own experience (Kember & Kwan, 2000). In contrast, studentcentred pedagogy has been described as shifting the teacher role from instructor to facilitator (Regan et al., 2012) whereby students discover and create knowledge (Kember, 1997). Student-centred pedagogy is reportedly a complex undertaking (Sadler, 2012) and should not be mistaken for students being expected to teach themselves; rather, it is a "balance between teacher-and student-directed learning" (Gordon, 2016(Gordon, , p. 1081.
Moving to online education, González (2009) found that the differences between teachercentred and student-centred online pedagogy can be seen in how learning management system (LMS) tools are utilised by teachers. He suggested that a feature of teacher-centred pedagogy online is providing structured learning materials to students for independent study. In contrast, student-centred online pedagogy focuses on providing students with communication spaces and opportunities to build knowledge. A defining feature of student-centred online pedagogy is reportedly the teacher facilitating collaboration between students for the purpose of knowledge discovery and creation (González, 2009(González, , 2010(González, , 2012. It has been found that teachers with teacher-centred beliefs will normally choose technological tools that let them teach that way, and in contrast, teachers with studentcentred beliefs will usually choose tools that support their enactment of studentcentred practices (Tondeur, van Braak, Ertmer, & Ottenbreit-Leftwich, 2017). Importantly, González (2013) found that teachers who align with student-centred conceptions of teaching but who perceive their online teaching context to be constrained will shift their practice to be more teacher-centred. In this way, enacted practice can represent a compromise between teaching conceptions and teaching context (Kennedy, 2010;Norton, Richardson, Hartley, Newstead, & Mayes, 2005). Critically, Lowenthal, Nyland, Jung, Dunlap, and Kepka (2019) found that an important factor in the effect of teaching context on practice is the teacher's own perceptions about the affordance or limitation of that teaching context. In this way, the relationships between conceptions of teaching, teaching context, and enacted practice remain dynamic.
Student-centred pedagogies are considered best-practice online, and according to Sun and Chen (2016), this includes encouraging students to discuss learning content with each other in relation to their own experiences. However, not all students value such discussions, and some perceive student-student communication does not contribute to their learning (Jaggars & Xu, 2013). Furthermore, the nature of student-centred pedagogy continues to have inconsistent interpretations and understandings (Trinidad, 2019). Therefore, understanding what a teacher conceives as student-centred pedagogy is important.
As described next, this study investigated the centredness of what a team of wholly online teachers conceive as good pedagogy, how aligned or not aligned those conceptions are with enacted practice, and factors within the teaching context that affect enacted practice.

Research context and design
In this section, the research context is introduced and then the design and analysis of the implemented digital survey is described.
In Australia, the vocational education and training (VET) sector is understood to be fundamentally different from other post-secondary teaching contexts (Fowler, 2017). VET is governed by principles of student competence (Harris & Hodge, 2009) and the very nature of this competency-based framework reportedly encourages teacher-centred pedagogy (Brennan, McFadden, & Law, 2001). This is because the VET curriculum is enforced through audits that require teachers to prove they have conveyed occupation-specific, curriculum-bound skills (Black & Reich, 2010). Importantly, transmission of skill is privileged over development of underpinning knowledge within VET (Wheelahan, 2009). Therefore, direct-instruction and transmission of information are encouraged pedagogies within VET, and Griffin and Mihelic (2019) have anticipated that this may introduce unique tensions when VET teachers are asked to teach online.
VET is a critical contributor to Australia's labour market (Wheelahan & Moodie, 2011) and is the country's largest education sector (Atkinson & Stanwick, 2016). More than half of all occupation qualifications in Australia need to be at VET rather than university level (Productivity Commission, 2011). Currently, 14-20% of VET units are taught online (NCVER, 2018(NCVER, , 2019Reeson, Mason, Sanderson, Bratanova, & Hajkowicz, 2016). However, for VET to continue meeting its responsibilities to skill and upskill current and future workers, it is shifting to expand its online education offerings (Reeson et al., 2016).
This study is situated within a wholly online VET teaching context. It profiled teachers as they move from conceptualisation to enacted online practice, identified how ideal and enacted practice were aligned or not aligned, and examined what teaching context factors affect enacted practice.

Participants
Teacher participants for this survey were sought from a single online VET teaching context so that results would not be affected by different institution teaching contexts. The largest VET provider in the state of Queensland, Australia, agreed to host the project and provide access to what is considered a large team of fully online teachers. That team included 66 teachers who teach more than 40,000 subjects a year online. After survey piloting, ethical approvals were obtained from both the researcher's university and the host site. Following principles of informed, voluntary consent, a personalised invitation and survey access token was sent to each target teacher. After cleaning up data to remove surveys less than half complete, 46 survey responses were retained representing a 69% useable response rate. Participants represented six teaching disciplines. The largest group of respondents came from early childhood (33%) followed by business (28%), and the characteristics of participants are presented in Table 1. They were generally female, full-time teachers, and aged 51 years old which is normal for the VET sector where teachers are thought to perceive the role as a suitable transition to retirement (Productivity Commission, 2011). The typical teacher in this study would be teaching 12 different online classes at any one moment in time, with a total of 131 students across those classes. Some classes have only one student. At time of survey, participants had an average of 4.5 years online teaching experience and will not have studied online pedagogy. Cox and Prestridge Research and Practice in Technology Enhanced Learning (2020)

Survey instrument
To investigate the pedagogy of online VET teachers, a digital survey was designed. The survey questions were developed from extensive reviews of online education and VET literatures conducted as part of a larger doctoral research project. The survey also incorporated previously validated survey tool components. The three investigative focuses for this survey were (1) what practices teachers conceive represent good online teaching, (2) is enacted practice aligned or not aligned with that ideal, and (3) what factors in this online VET teaching context affect practice. The survey commenced with background and demographic questions to establish who VET online teachers are, their teaching experience, and the nature of their teaching role such as numbers of students. Teaching profiles were then built classifying conceptions of teaching and enacted practice along a continuum from teacher-centred to student-centred. Two different approaches were utilised to yield teaching profiles.
The first approach incorporated a validated survey tool that was obtained from Owens (2015) and then adapted with permission. Those adaptations included changing some language from the higher education sector in the UK to language more familiar to VET teachers in Australia (see the Appendix for adaptations made). Owens' 5-point scale disagree-agree questions incorporated topics posed from student-centred and teacher-centred perspectives. Those topics were posed twice. The first question set was designed to yield a composite result for each teacher to represent their conception of teaching profile from teacher-centred to student-centred. The composite result from the second question set then yielded a profile of enacted practice, thereby enabling comparisons between conceptions of teaching and enacted practice for each teacher and the group. By design, these conceptions of teaching and enacted practice question sets were presented within the online survey on different screens and separated by unrelated questions to reduce participant teachers consciously linking them.
The second approach to profile teaching in this context drew on Owens' design of directly comparing teaching conceptions and enacted practice. The questions for this second approach were original to this survey and were developed from a model purported to represent best-practice student-centred online pedagogy. That model was Bain's description of what the best teachers do (Bain, 2004) as updated for online education by Brinthaupt et al. (2011). Nine student-centred practices drawn from that work (see the Appendix) were presented to teachers twice. Firstly, using a 5-point scale, teachers were asked to rate the importance of each practice. Secondly, teachers were asked to report the frequency with which they enacted each practice. Those two question sets were presented in the survey on different screens to reduce participant awareness of their relatedness. Results from the two approaches were utilised to develop teaching profiles and enabled differences, similarities, and nuances between conceptions of teaching and enacted practice were identified.
Separately within the survey, teachers were asked whether particular teaching context factors affected their practice (see the Appendix). Given the lack of literature about online VET, those items were developed from a detailed review of literatures about higher education online teaching contexts and from face-to-face VET teaching contexts.

Analyses
Non-parametric tests were utilised in the analysis because they do not assume the sample is normally distributed (Field, 2013) and are therefore ideal for small sample sizes, especially with categorical and ordinal data (Pallant, 2016). The Wilcoxon Signed Rank Test was utilised to compare participants under two different conditions (Field, 2013;Pallant, 2016), such as when comparing conceptions of teaching and enacted practice responses. The results of this test were reviewed adjacent to Friedman Tests to confirm consistency of result. The magnitude of the effect was interpreted using Cohen (1998) as cited in Pallant (2016): 0.1 = small effect, 0.3 = medium effect, 0.5 = large effect.
Separately, the Kruskal-Wallis H Test is recommended to be suitable for comparing differences between more than two groups by using ranks (Field, 2013;Pallant, 2016). It was utilised in this analysis when comparing participants by group such as by age, years teaching experience, or discipline being taught. Such analysis identified if, for example, participants with higher level teaching qualifications had similar answers within a question or question set. This paper reports only those results relevant to the topic and that yielded a significant result, seeking < 0.05.
The internal consistency of the survey items were tested using Cronbach's alpha coefficient (Pallant, 2016). Scales reported in this paper were deemed reliable and have alpha coefficients of between 0.67 and 0.91. However, an acknowledged limitation of this study is its small sample size. The survey has internal validity in that it accurately measures what it was intended to measure (Cohen, Manion, & Morrison, 2011;Creswell, 2011). However, with only 46 useable responses, the findings are informative in nature and not indicated as generalisable or transferable across teaching contexts.

Results and discussion
In this section, survey results are presented and discussed in three parts: (1) teacher characteristics, (2) teaching profiles using two different approaches, and (3) relationships between teaching context and enacted practice.

Teacher characteristics
The teacher's age, gender, seniority of teaching role held, level teaching and industry qualifications completed, or years teaching and industry experience were not found related to enacted practice. However, having been an online student was found related to enacted practice. It has been reported in literature that many teachers have not been an online student (Niess, 2011). The teachers in this sample are different in that 71.7% reported they have been an online student. Breaking this down, 39.1% of teachers had previously completed a course online and 32.6% were currently completing a course online. This is notable because teachers in this study who had been an online student reported more frequently enacting the student-centred practices of creating a friendly, social learning environment (H(1) = 4.02, p = 0.045), developing group cohesiveness among students (H(1) = 4.36, p = 0.037), and consciously building rapport with each student (H(1) = 10.10, p = 0.001) than teachers who had not been an online student.
While teaching qualifications were not found related to enacted practice, other relationships were identified. Literature has reported that the mandated Certificate IV Training & Assessment is likely the highest-level teaching qualification held by VET teachers in Australia (Smith, Hodge, & Yasukawa, 2015). However, in this study, only 32.6% of teachers reported this certificate as their teaching qualification with the remaining teachers holding higher level teaching qualifications than required. Notably, 41.3% of participant teachers reported holding a university-level teaching qualification, and 21.7% of participants were currently undertaking study to obtain university-level teaching qualifications. Interestingly, 15.2% of participant teachers were currently studying a university qualification as an online student. The level of teaching qualification held is relevant because teachers without university-level teaching qualifications were found to more likely (Md = 27) want additional training for the development of pedagogical knowledge than teachers with university-level teaching qualifications (Md = 17). H(1) = 7.68, p = 0.006. However, it is important to note that regardless of teaching qualification held or currently being pursued, only 15.2% of teachers had studied online pedagogy as part of their qualification.
Next, teacher profiles are reported.

Teaching profiles
This section presents teaching profiles that were captured in the survey through two different approaches. Within each approach, the centredness of teaching conceptions are profiled along the teacher-centred to student-centred continuum and then compared to centredness of enacted practice.
Teaching profiles using approach one As presented in This revealed that an intermediate answer option was only selected by teachers for 14% of all responses. Instead, most teachers selected teacher-centred answers for some questions and student-centred answers for others. When summated, this resulted in an unexpectedly high number of intermediate profiles. That outcome was the result of Owens' incorporation of two questions types-some with student-centred language and others with teacher-centred language. When sorted from highest to lowest mean using Owens' topic groupings (Table 3), a pattern emerged whereby student-centred questions had yielded 89.1% student-centred responses, and teacher-centred questions had yielded 78.3% teacher-centred responses.
Some questions were reversed when posed to teachers, so having the participants simply agreeing with all statements was not the cause of the patterns within teachercentred and student-centred questions. To explain the responses, three teacher-centred question topics are now unpacked.
Two questions were posed to teachers about the purpose of education within the topic training for jobs. The two questions were as follows: The main aim of teaching should be job specific training and An important function of [VET] education is to produce graduates for specific jobs. The publicly stated purpose of VET is to produce work-ready students to meet labour market demands (Productivity Commission, 2019), and therefore, it is not surprising that teachers agreed with this statement about the purpose of the education they provide. Following Owens' design, that agreement resulted in teacher-centred profiles for those questions.
Two questions were posed to teachers about imparting information. These two questions were as follows: A good online teacher is one whose main role is to impart information to his/her students and Teaching is about the transmission of content or skills. The nature of competency-based VET curriculum assumes that skills can be divided into isolated pieces which do not require underpinning or relational knowledge (Wheelahan, 2005) and as such, within VET job tasks are prioritised over knowledge (Wheelahan, 2009). Therefore, the VET teachers having agreed to these statements makes sense. Again, following Owens' design interprets that agreement as teachercentred.  Finally, a question was posed about knowledge of subject. That question was It is fundamental that online teachers know the latest knowledge and skills related to their subject area. Importantly, the main requirement to become a VET teacher is deep industry experience (Kemmis & Green, 2013;Kemmis, Hodge, & Bowden, 2014) which is privileged over teaching skills or knowledge (Simons, Harris, Pudney, & Clayton, 2008). Agreement with this sentiment fits the teaching context in which it was being answered, and resulted in teacher-centred profiles for this question.
The responses for these topics are likely embedded in the nature of VET teaching context. This highlights differences between VET and higher education where the questions were first developed by Owens. Separately, within the student-centred topics, the question set with the lowest-mean was the topic interactive teaching which relates to student-to-student interaction. This indicated that student-student practices were perceived as less important than teacher-centred practices.
The skewed results between the teacher-centred and student-centred question sets meant that using the composite results and Kruskal-Wallis H Tests to identify a relationship with enacted practice was not feasible. However, as presented in Table 4, Wilcoxon signed-rank tests, confirmed with Friedman tests, revealed statistically significant differences between eight of 16 paired conceptions of teaching and enacted practice questions sets. An important finding from these results is that enacted practice was less student-centred than conceptions of teaching for 13 of 16 question pairs. The differences between conceptions of teaching and enacted practice were significant for eight of 16 question pairs.
In summary, although not yielding a clear student-centred or teacher-centred conception of teaching profile, this approach revealed that enacted practice was less student-centred than conceptions of teaching for 13 of 16 question pairs. Notably, the nature of VET was found to affect the teaching conceptions of its teachers.
A different approach to finding teaching profiles is reported next.

Teacher profiles using approach two
This approach found that 71.7% of teachers profiled as having a student-centred conception of teaching, 19.6% as intermediate, and 8.7% as teacher-centred (Table 5). Different to the strategy reported in the previous section, this second approach resulted in consistent responses across all nine questions. When examining individual questions to understand what a student-centred profile meant, it was identified that the three questions with the lowest means were the three questions related to student-tostudent collaborative practices. This is an important observation because, as established previously, an identifying feature of student-centred pedagogy online is student-tostudent collaboration. Table 6 presents conception of teaching responses sorted highest to lowest mean.
In contrast to the 71.7% of teachers profiled with student-centred conceptions of teaching, only 23.9% profiled as also enacting student-centred practice. Instead, 45.7% of teachers enacted intermediate practice and 30.4% teacher-centred. Teachers reported enacting teacher-centred practice for five of nine practices in this set.
Relationships between conceptions of teaching and enacted practice were found. Teachers with a student-centred conception of teaching reported a higher frequency Cox and Prestridge Research and Practice in Technology Enhanced Learning (2020)    (Md = 27) of enacting the practice of using discussion forums to facilitate student-tostudent social interaction than teachers with an intermediate (Md = 14) or teachercentred (Md = 17) profile (H(2) = 7.78, p = 0.02). Teachers with a student-centred conception of teaching also reported a higher frequency (Md = 26) of enacting the practice of creating a friendly, social learning environment than teachers with an intermediate (Md = 15) or teacher-centred (Md = 19) profile (H(2) = 6.30, p = 0.04). Finally, teachers with a student-centred conception of teaching reported a stronger desire (Md = 25) for more training about how to effectively utilise technology to teach their discipline than teachers with an intermediate (Md = 16) or teacher-centred (Md = 14) profile (H(2) = 7.42, p = 0.02). An important finding from this part of the survey was that all nine enacted practices were found to be less student-centred than their conception of teaching counterpart. Those differences were statistically significant for seven of nine pairs (Table 7).
In summary, using this second approach, conceptions of teaching were profiled as being student-centred and relationships between conceptions of teaching and enacted practice were found. Importantly, enacted practice was less student-centred than teachers' ideal, and statistically significant differences between teaching conceptions and enacted practice were identified.
Drawing on both survey approaches identified teachers conceived that studentcentred practices represent good pedagogy online. Notably student-to-student collaborative learning was less important than teacher-student practices, and the context of VET resulted in teacher-centred conceptions regarding the nature and purpose of VET education. Furthermore, enacted practice was less student-centred than teachers perceived as ideal. Many of the shifts between conception and practice were statistically significant. Enacted practice being different from ideal has been previously suggested as related to the teaching context which can also affect enacted practice. Findings related to teaching context are presented next.

Teaching context
To understand teaching context effects on enacted practice in this online VET teaching context, eight factors were identified from literature; four of these were included directly in the survey through targeted questions, and four were investigated through relationships between variables.
Teacher workload has been identified in other teaching contexts to affect enacted practice (Conrad, 2012;Murphy & Rodriguez-Manzanares, 2012). That was supported in this survey where teacher workload was reported by 66.7% of teachers as often or always preventing them from enacting a teaching practice they believe would be beneficial. As illustrated in Table 8, which is presented highest to lowest mean for actual workload, teachers further reported that their two highest workload tasks were associated with assessing competencymarking assessments followed by helping students complete their assessments. Notably, only 31.0% of teachers reported wanting to spend less time on these tasks. The third highest workload task was administration and 61.9% of teachers would like to spend less time on this. Teachers would like to spend more time actively facilitating student learning.
Two teaching context factors that may be related to workload were investigated through analysis of relationships between variables. These are the number of classes and students per teacher. Teachers in this study reported being responsible for teaching between two and 57 online classes at any one time (x = 12). However, enacted practice was not found to be affected by the number of classes per teacher. Furthermore, no relationship between the number of classes and workload responses was found. Different to the number of classes per teacher was the number of students they were each responsible for teaching. The number of students per teacher ranged from 10 to 310 students (x = 131), and the number of students per teacher was found to have an effect on one conception of teaching variable and one enacted practice variable. Teachers with > 175 students were more likely than teachers with less students to agree that students should be helped to learn for themselves (H(2) = 7.70, p = 0.02). Teachers with < 75 students were more likely than teachers with higher student numbers to report more frequently enacting the student-centred practice of using technology to engage with students in real-time (H(2) = 7.55, p = 0.02). Next, the effect of class size on enacted practice was investigated because, as reported by Cox, Black, Heney, and Keith (2008), "a teacher with over a hundred students will by necessity make different choices than one with only twenty students" (p. 380). Results from this survey supported and built on that assertion. Based on feedback from the institution prior to survey implementation, small class sizes were defined in this study as less than five students and large class sizes were defined as more than 75 students. Using those definitions, 52.6% of teachers reported that small class sizes do affect enacted practice. However, the reported nature of that effect was a mix of positive (31.6%), negative (7.9%), or both positive and negative (13.2%). In contrast, large class sizes were reported by 58.8% of teachers as negatively affecting their enacted practice, and 17.6% of teachers reported that large classes both positively and negatively affect enacted practice.
Course duration has been previously indicated by Akyol, Vaughan, and Garrison (2011) as an influence on enacted practice in a higher-education online teaching context. However, a relationship between course duration and enacted practice was not found here. 51.4% of teachers in this study reported that they did not perceive their 12week course duration affected their enacted practice, and 35.1% of teachers perceived course duration as a positive influence. The discipline being taught affecting enacted practice has been suggested in highereducation online teaching contexts, and Arbaugh, Bangert, and Cleveland-Innes (2010) proposed that mathematics teachers, for example, may not attempt to enact studentcentred practices. As suggested by Hornik, Saunders, Li, Moskal, and Dzuiban (2008), some teachers may perceive collaborative pedagogy is not an appropriate online strategy when teaching linear, fact-driven concepts. Some relationship between the discipline being taught and enacted practice was indicated here. Teachers in this study represented the disciplines of early childhood, business, accounting, education and library (schoolteacher aides), justice and government, and information technology (I.T.). Analysis revealed that I.T. teachers were more likely to report a higher frequency of using discussion forums to facilitate student-student social interaction (H(5) = 11.26, p = 0.04), of using technology for real-time student engagement (H(5) = 22.00, p = 0.00), and for getting students to participate in online discussions (H(5) = 12.42, p = 0.03). Justice and government teachers were more likely to report a higher frequency of facilitating discussion forums for students to explore concepts and develop deep knowledge together (H(5) = 12.63, p = 0.03). Accounting teachers were more likely to report a higher frequency of ensuring students are well skilled in the subject competencies (H(5) = 17.16, p = 0.00). However, given the small size of this study, generalisations cannot yet be drawn although these indicators warrant scaling the survey to a broader population.
It has previously been reported that compliance with Australia's federally mandated VET curriculum affects enacted practice for VET teachers who are expected to interpret that curriculum and make pedagogical decisions on the basis of that interpretation (Hodge, 2014). VET teachers must then formally demonstrate their compliance with that curriculum in how they teach and then assess student competency (Black & Reich, 2010). Relationships between VET curriculum and enacted practice were found from this survey. Teachers reported that complying with the mandated curriculum sometimes (51.2%) or often (14.6%) prevented them from enacting a teaching practice they believed would be beneficial for their students. Teachers further reported that complying with curriculum sometimes (46.5%) prevented teaching a concept they believed important for their students to learn. Notably, administration associated with proving curriculum compliance was ninth highest of 15 workload tasks, and 31.0% of teachers would like to spend less time on this. Compliance-related workload being time consuming and leading to a compromise in what teachers perceive as quality enacted practice has been reported previously for face-to-face VET teaching contexts (Black & Reich, 2010). More recently, Schmidt (2019) found that non-teaching workload tasks affect enacted VET practice through less time being available to teach.
Finally, literature has previously reported that professional development is lacking for VET teachers and is not always relevant for online teachers. Those perceptions were not supported by this group of online VET teachers who agreed or strongly agreed that professional development was regularly available (73.9%), that it was relevant to their online teaching role (54.3.%), and that it was high quality (56.5%). Enacted practice was not found to be affected by availability, relevance, and quality of professional development in this context.
In summary, of the eight teaching context factors investigated in this survey, five were found to affect enacted practice. These are teacher workload, small and large class sizes, compliance with the competency-based curriculum, the number of students per teacher, and the discipline being taught. The three investigated teaching context factors not found here to affect enacted practice are the number of classes per teacher, the availability, relevance, and quality of professional development, and course duration.

Conclusion
In Australia, VET is the largest education sector (Atkinson & Stanwick, 2016) and is an essential contributor to Australia's labour market (Wheelahan & Moodie, 2011). In recent years, 14-20% of all VET units have been taught online (NCVER, 2018(NCVER, , 2019Reeson et al., 2016), and the continued expansion of VET requires an expansion of its online education (Reeson et al., 2016). However, it has been suggested that VET may not be well suited to online education (Griffin & Mihelic, 2019). VET competencybased underpinnings are understood to foster teacher-centred pedagogy (Brennan et al., 2001), and the nature of VET may introduce tensions when brought together with online education (Griffin & Mihelic, 2019) where student-centred practices are understood to be the mark of good online pedagogy (Sun & Chen, 2016). Like other education sectors, VET currently experiences lower successful completion rates for its online students than its face-to-face students (DET, 2016;Griffin & Mihelic, 2019). Despite the importance of VET as an education sector, and despite its move toward online education, VET remains an under-valued research sector (Waters, Simon, Simons, Davids, & Harreveld, 2015) and little is yet understood about the pedagogy of VET online.
The purpose of this small study was to contribute an understanding of pedagogy within the combined yet potentially disparate teaching contexts of online education and VET. The implemented digital survey profiled what VET teachers conceived as good online practice, how ideal and enacted practice was aligned or not aligned, and what teaching context factors affect enacted practice. This resulted in three clear outcomes.
The first outcome was the finding that online VET teachers conceived good online pedagogy as student-centred. However, two complexities within teaching profiles emerged when analysing individual questions. Firstly, student-to-student collaborative learning practices were consistently rated lower than other student-centred practices. This is notable because one defining feature of student-centred online pedagogy is the facilitation of collaborative learning opportunities (González, 2009(González, , 2010(González, , 2012. The second complexity is that teachers selected teacher-centred responses against three particular topics. Their responses reflected that teachers conceived the purpose of VET education is job-specific training rather than content knowledge, that the role of teachers is to impart their industry-specific skills to students, and that it is important for teachers to maintain strong subject knowledge. This supports previous assertions that the nature of VET encourages some teacher-centred conceptions, and supports the need for research that is specific to VET teaching contexts. Cox and Prestridge Research and Practice in Technology Enhanced Learning (2020)  The second outcome from these survey results was that conceptions of teaching and enacted practice were not aligned, and those differences were statistically significant. Enacted practice was consistently more teacher-centred throughout the survey than teachers conceived was ideal. In particular, student-centred practices reported by teachers to be rarely or never enacted by them included teachers humanising themselves, using discussion forums for students to get to know each other and\or for building knowledge, utilising technology for real-time engagement, and developing group cohesiveness enabling students to work together for mutual benefit. Reasons for this gap between ideal and enacted practice likely include teaching context factors that, as described next, were found to affect enacted practice.
The third outcome was the identification of some teaching context factors that affect enacted practice. These factors were teacher workload, small and large class sizes, compliance with the competency-based curriculum, the number of students per teacher, and the discipline being taught. Teachers reported that workload, for example, prevented them from enacting practices they perceived would be beneficial for student learning. Furthermore, they reported that their highest workload tasks were marking assessments and reviewing draft assessments rather than other teaching practices.
This study contributes new knowledge about the pedagogy of online VET teachers and indicators about what influences their enacted practice. It has been contextualised within a large yet under-researched education sector, and results indicate a need for further VET-specific research. The results from this study will be used as a foundation to inform qualitative research to capture deeper teacher perceptions about their knowledge of online pedagogy, reasons for the differences between conceptions of teaching and enacted practice, and whether student-centred practices are encouraged or supported at their institution. Results from that subsequent research will be reported in due course.
The VET institution who provided access to online VET teachers for this study are navigating both the continued expansion of online education and the increased demand for its enhanced quality and effectiveness. That institution intends to use the results of this study to review their online teacher hiring, training, and management practices, and to better understand online education as more than simply a different mode of delivery. Their intent is that understanding and responding to factors that affect practice will support the enactment of online education that aligns to what is currently established in the literature as good online pedagogy for teachers, thereby subsequently enhancing student learning outcomes. An opportunity exists to investigate links between enacted practice and student outcomes within online VET.
For other researchers, although the small sample size within this study means the findings are not generalisable across other online teaching contexts, the survey is scalable. Broadening the application of this survey would contribute more to our understanding of what teachers conceive as good practice online and what shapes pedagogy as teachers in different online teaching contexts as teachers move from conceptualisation to enactment.
As a final note, this research was conducted before the 2020 pandemic. The shift to online education has moved more rapidly this year than ever before and unprecedented numbers of educators are currently navigating new online teaching contexts. These shifts make research such as this study all the more relevant because of the lessons that can be learned from the perceptions and practices of educators working in established online teaching contexts.  14. An important function of higher education is to produce graduates for certain professions within the community.
1.2 The main aim of teaching should be job specific training. 1.14 An important function of TAFE education is to produce graduates for specific jobs.
36. I use online environments to ensure that by the end of their course my students will be well qualified in their particular subject. 41. I use the online environment to prepare students for the roles they will have when they leave the institution.
2.14 I prepare online students for the roles they will have when they leave TAFE. 2.9 I ensure that by the end of their course my students are well skilled in the unit competencies. Imparting information 4. A good lecturer is one whose main role is to impart information to his/her students. 6. Teaching is about the transmission of knowledge.
1.4 A good online teacher is one whose main role is to impart information to his/her students. 1.6 Teaching is about the transmission of content or skills.
34. I use the online environments to pass on what information I know to students. 39. Within the online environment I give as much information as possible to my students.

I pass on what
information I know about the topic to students. 2.12 I try to give as much information as possible to my online students.
Knowledge of subject 11. It is fundamental that lecturers know the latest advances in knowledge related to their subject area. 16. A good lecturer has to be an expert in their subject matter.

Table 11
Approach two for identifying conceptions of teaching and enacted practice. Nine student-centred indicators developed from Bain's 'what the best teachers do' (Bain, 2004) adapted to online education by Brinthaupt et al. (2011) Category Question prompt

Stimulate intellectual development
Utilise engaging tasks students will find interesting. Utilise a variety of technologies such as videos or podcasts. Facilitate discussion forums where students explore concepts and develop deep knowledge together. Utilise technology for real-time engagement with groups of students (e.g. Live Rooms) Foster student engagement Develop group cohesiveness, helping students work together for mutual benefit. Use discussion forums to facilitate social interaction between students. Create a friendly, social atmosphere where deep learning is encouraged.
Build rapport with students Use introductory videos or other self-disclosure methods to humanise yourself to students. Consciously build rapport with each individual student.

Cox and Prestridge
Research and Practice in Technology Enhanced Learning (2020) 15:16