The case for engaging online tutors for supporting learners in higher education in refugee contexts

Research in Learning Technology 2020. © 2020 P. O’Keeffe. Research in Learning Technology is the journal of the Association for Learning Technology (ALT), a UK-based professional and scholarly society and membership organisation. ALT is registered charity number 1063519. http://www.alt.ac.uk/. This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.


Introduction
The global refugee population is at its highest point in history. With an estimated population of 65.6 million people forcibly displaced around the world (UNHCR 2017), the international community is being pressed to find solutions to this ever-growing crisis. Beyond the challenge of providing food, shelter and safety for those that it does so far, the international humanitarian and development communities are increasingly being called on, by various initiatives such as the Sustainable Development Goals (SDGs) 1 , to provide other more long-term solutions to improve 2 Citation: Research in Learning Technology 2020, 28: 2428 -http://dx.doi.org/10.25304/rlt.v28.2428 (page number not for citation purpose) the lives of the millions affected by global migration. Education, and increasingly higher education, is at the forefront of such initiatives. With just 1% of the global refugee population attending higher education (compared with 34% of the global non-refugee population) (UNHCR 2015), there is scope for higher education interventions to make an impact on providing some pathways out of the global refugee crisis.
Traditional higher education responses, such as increasing university seats for refugees and providing scholarships, although welcomed and useful for some, require infrastructure that many of the world's refugees cannot access. The remote location of refugee camps and the lack of safety and security for students to attend classrooms in conflict situations means that the bricks and mortar approach to providing higher education solutions for refugees is not always an appropriate response (Dahya 2016). Instituting innovative responses that harness the potential of information communication technology (ICT) to deliver higher education is one area that circumvents the limitations of more traditional 'college campus' responses. Distance learning programmes that utilise ICTs such as massive open online courses (MOOCs) and blended learning initiatives are helping to bring higher education opportunities to refugees confined to remote camps (physically or financially) and offer pathways to brighter futures for them and their communities.
ICT responses have gained a great amount of traction in recent years in development circles as they not only potentially provide education to more people in more parts of the world but also cost less than traditional education responses (Balkin and Sonnevend 2016). Through the proliferation of mobile devices and internet access among refugee populations, the innovation of MOOCs has been hailed as a model for facilitating the democratisation of higher education among refugee communities (Greenaway et al. 2016). However, MOOCs tend to be designed with the needs of learners in developed countries in mind and frequently try to simulate the live classroom experience through delivery methods such as the use of lecture streaming (Moser-Mercer 2014). Furthermore, it has long been recognised that education needs more than a classroom and, as such, providing higher education is not just an issue of providing the access that MOOCs do. To be successful, the delivery of education in fragile contexts requires more than mere access, with a fortified system that provides adequate support ranging from infrastructure to administration (Dahya 2016).
A sufficient learning-enabling experience goes beyond granting access to online materials and encompasses an interconnected web of support that enables effective education engagement to allow for the effective transfer, processing and acquisition of concepts and knowledge. In this study, scaffolding that comprises InZone's learning ecosystem for enabling higher education in refugee contexts is presented as a case study to illustrate how one such higher education in refugee context programmes works. While looking at the overall learning ecosystem, the role of one of its key actors, the online tutor, is focused on with the aim of better understanding the pedagogical role that online tutors play when enabling a higher education experience for learners in refugee contexts. By examining the role of the online tutor in the context of InZone's learning ecosystem, the author aims to lay out the tutor's function and explore the tutor's usefulness in assisting refugee learners by presenting how these challenges were approached and mitigated using the InZone tutoring programme.

ICTs, collaborative learning and the role of the online tutor
Tutoring has a long history as an educational support mechanism stretching as far back as educational practices have been recorded (Hartman 1990). Through the ages, it has taken various forms from peer-to-peer tutoring in class to paraprofessional instruction simulating real-life 'on-the-job' scenarios (Cohen et al. 1982). With the increasing importance of ICTs in the educational sphere, it follows that tutoring has evolved along with the development and incorporation of ICTs into the education process. As course delivery overcomes the traditional spatial and temporal limitations that brick and mortar universities impose, the use of ICTs for education delivery enables new forms of tutoring that overcome the geographical confines of the past. Not only do ICTs allow education connections to be made traversing the traditional limitations of time and space, but ICTs also allow education opportunities to reach many more people than they previously could. It is now possible to bring higher education to far-flung reaches of the world and, as such, involving higher education in the development mix for refugee communities has become a more viable option than ever before.
Although in its vanguard, research and literature assessing the role that ICTs can play in providing higher education opportunities in refugee contexts is underdeveloped and sparse (Dahya 2016). The relatively new nature of this field, coupled with the difficulty in conducting research in volatile contexts, has led to a thin pool of resources, largely confined to conflict and development studies. Despite the lack of study in this area, the research carried out to date suggests that merely connecting students to teachers thousands of miles away from each other across an ICT platform becomes futile unless various supports are in place (Moser-Mercer 2014; Moser-Mercer, Kherbiche, and Class 2014). Focusing on the support role ICT can play in delivering education in fragile contexts, Dr Neghin Dahya's 2016 landscape review Education in Conflict and Crisis -How Can Technology Make a Difference? 2 provides a general overview of the research available in conjunction with interviews carried out with 15 expert practitioners in the field. In terms of literature relating directly to the role of online tutors in supporting refugees in fragile contexts, research is even thinner on the ground. As a result, to gain an appreciation of the role online tutors play in enabling higher education for refugee learners in emergencies and fragile contexts, it is necessary to look at the role tutors play in scaffolding online education and to try to relate it to the refugee context.

Locating the online tutor in the collaborative learning ecosystem
Before analysing and assessing the role online tutors play in delivering higher education in refugee contexts, InZone's 3 collaborative learning ecosystem for enabling higher education refugee contexts is presented as a model to understand this form of pedagogy. This ecosystem has been developed through InZone's decade long experience of enabling higher education for refugee learners in refugee camps in the Middle East and Africa.
Acknowledging the necessity for support in ICT facilitated learning as outlined in the introduction, Collaborative Learning Theory offers a framework from which a learning ecosystem can utilise various support mechanisms to deliver facilitated learning. The theory, at its most basic level, can be said to be a situation where two or more people learn together, or attempt to learn together (Dillenbourg 1999). It is borne out of the Vygotskian perspective, which sees learning as essentially a sociocultural process that understands cognitive development as requiring interaction between two or more people (Lantolf and Pavlenko 1995;Lantolf and Thorne 2006). Learning from this theoretical perspective is not just a matter of peers interacting and learning from each other, but can also require individuals interacting with mentors and more knowledgeable peers who enable learning through a process of guidance and/or collaboration (Lin 2015). The InZone collaborative learning ecosystem embraces this perspective by placing the refugee learner in an ecosystem where he or she interacts with his or her peers, as well as with various other actors in the learning ecosystem who facilitate and collaborate in the learning process.
With a decade-long experience of observing and teaching refugee learners in refugee camps and extensive knowledge of enabling blended learning courses in refugee contexts, the InZone collaborative learning ecosystem model has been empirically developed to accommodate five key actors -students, lecturer, course coordinator, on-site facilitator and online tutor. The key actors work together in the ecosystem to maximise knowledge transmission, processing and acquisition and foster an increasingly autonomous culture of learning among learners. The following diagram illustrates the actors and their inter-relationships in the InZone collaborative learning ecosystem. This is followed by a brief explanation of each actor and his or her role before exploring the role of the online tutor in more detail.

(page number not for citation purpose)
InZone's collaborative learning ecosystem is populated by five principal actors: lecturer, on-site facilitator, online tutor, course coordinator and the student. The following explanations elaborate on the role of each of these actors: 1. The lecturer delivers the course material online, encourages the generation of new knowledge and evaluates the student's learning. In the ecosystem, the delivery of knowledge via an online platform enables the transmission of information to the students, who through discussions, group work, interactions and so on acquire and develop new knowledge. 2. The on-site facilitator provides on-site technical and guiding support to learners, helping them to access the learning platform on location and navigate the physical learning space. The on-site facilitator is a critical contact point in the education relationship between the students and the other members of the collaborative learning ecosystem as they meet the students on a frequent basis in the learning space, as well as interact with them on the ICT platform. (WhatsApp is primarily used by InZone as it is the easiest and most available for students, tutors and facilitators.) 3. The online tutor is an online subject matter expert or a peer with an advanced level of subject knowledge. The tutor plays a pedagogical role in this collaborative learning ecosystem by 'meeting' the student regularly over an ICT platform (WhatsApp) to stimulate new knowledge acquisition, discuss the student's progress and offer advice on becoming a successful independent learner. 4. The course coordinator has the overall responsibility of ensuring day-to-day running of the course and liaises with the other members of the learning ecosystem to ensure a smooth operation. 5. The student is the focal point of the learning ecosystem. This means that they are central to the collaborative learning model and the entire learning ecosystem is designed to support their optimal learning by meeting their educational needs and promoting progressive learner autonomy.
While all actors in the collaborative learning ecosystem are necessary for the success of the model, the primary focus of this study is on the role of the online tutor. As such, the rest of this paper will concentrate primarily on exploring this role, which is contingent on the other actors in the ecosystem operating efficiently in the model.

Online tutors and peer-to-peer learning in the collaborative learning model
Classroom talk has long been recognised as a central component of effective learning (Dooley 2009). Similarly, peer-to-peer learning has also been recognised for its prowess and vitality in knowledge generation, acquisition and processing. In a nod to their constructivist roots, modern education theories hinge education attainment not just on the relaying of knowledge from teachers to students but also on the collaborative interaction between students and their peers. As the name implies, collaborative learning involves two or more people working together in pairs or in small groups to discuss concepts and find solutions to problems (CTE 2017). Collaborative learning is credited with better retention, equipping students with more 'real world' problem-solving knowledge and skills and enhancing critical thinking skills (Gokhale 1995). Academic peer tutoring can take many different forms, from surrogate teaching to reciprocal peer tutoring (Briggs 2013). In InZone's collaborative learning ecosystem, online tutoring follows the concept where the tutor is more advanced in terms of subject matter knowledge and the student-tutor relationship usually involves one tutor to a small group of students. For example, InZone uses the model for enabling a basic medical training course at Kakuma Refugee Camp in Kenya that connects refugee learners with medical student tutors from the University of Geneva. The University of Geneva tutors are more advanced in their medical training than the refugee learners and 'meet' regularly via the WhatsApp instant messaging platform to discuss new topics, acquire new knowledge and encourage progressive learning autonomy.
While the expectations placed on the online tutor differ according to the specifics of each of InZone's courses and the demands of the students, the general rule of thumb is that the online tutor plugs into the collaborative learning model by enabling the peer-to-peer learning element necessary for a successful educational experience. For example, the medical tutors pose a question on the platform. All the students and the tutors discuss this question in a collaborative learning manner, allowing the group to learn together and acquire deeper knowledge.
Online tutors are put forward as candidates to InZone by course leaders in advance of a course beginning. Each tutor then goes through an intensive 'online tutoring in refugee contexts' training programme, which has been empirically developed by InZone. This involves taking part in face-to-face training, an online training course, simulation and guidance through the initial tutoring process. The tutoring sessions take place at least once a week over an ICT platform and follows a regular pedagogical plan for online teaching as laid out by the course leader in the course syllabus. A typical online tutoring session aims to compliment online class learning and involves two tutors 'meeting' the students over WhatsApp to discuss a particular learning point and stimulate collaborative learning.
At the most basic level, collaborative learning is not an obvious component of online courses. By their very nature, online courses tend to be more individually focused and involve solitary pursuits when compared with traditional educational experiences -thus not fitting well with the sociocultural perspective of cognitive development. The spatial and temporal limitations that MOOCs and other online courses overcome may, on the face of it, come at the price of limiting or even excluding the role peers play in knowledge production and transfer. This potential collaborative learning deficit appears to have been recognised quite early on in the development of MOOCs. In fact, the precursors to MOOCs, correspondence courses, set the scene for this as they were heavily criticised for lacking the key pedagogical component of peer interaction (Marques 2013). Since their inception, extensive research has been carried out into the development of computer-supported collaborative learning (CSCL) technologies (Manathunga and Hernández-Leo 2015) and the incorporation of social media support (e.g. Facebook groups) as a means to bring peer-to-peer elements into digital learning delivery (Balkin and Sonnevend 2016). While it is beyond the scope of this study to assess the role of CSCL and social media in education delivery, it is worth noting that there is widespread recognition among course providers that collaborative learning is a vital component and online courses that omit it may operate on a deficit. 7 (page number not for citation purpose)

Key factors that influence online tutoring
Peer-to-peer learning, as noted, is a long-recognised component of modern-day pedagogical practice and vital for collaborative learning. Extensive research supports the positive learning impact of peer-to-peer tutoring in face-to-face contexts (De Smet et al. 2010). Similarly, although less extensively due to its relative novelty, research tends to support the positive impact online peer-to-peer tutoring has on learning (De Smet et al. 2008;McLuckie and Topping 2004). Peer-to-peer tutoring, without question, plays a positive role in the generation, acquisition and processing of concepts and knowledge. What is less understood, however, is the impact that individual tutor behaviour plays on the learning process and how cross-cultural divisions between western-based tutors and refugee learners located in emergency and fragile contexts affect this behaviour. In addition, there is ample evidence to suggest that online communication differs greatly from face-to-face communication (Okdie et al. 2011) and thus may affect the learning process.
When a tutor is recruited to tutor an InZone course, they receive intensive training on how to deliver online tutoring. This training takes place in advance of a course and involves a 2-h, face-to-face, structured training seminar and an online training course 4 tailored to the needs of the tutors, which develops the tutors' abilities to teach online and in refugee contexts. The training has been developed to address the three central issues of individual behaviour, cross-cultural differences and online communication, which affect the online tutoring experience. Training support is ongoing throughout the tutors' tenure with InZone, with the course coordinator monitoring weekly tutoring sessions and additional training being given as needed.
The subsequent sections look at how these three factors can influence online tutoring of refugees and highlight how their impact is mitigated through InZone's programme.

Tutor behaviour
The behaviour of peer tutors, while clearly having an impact on the learning process, tends towards more directive and knowledge-telling modalities rather than facilitative and constructivist-oriented strategies (Berghmans et al. 2012). This may be a function of the process where the tutor is a subject matter expert and the student a novice and thus may affect the move from transmissive to collaborative learning. Online tutoring, where the tutor is a subject matter expert, may also carry these assumptions. In trying to shift from a transmissive to a more collaborative learning model, such behaviour may be an issue. As subject matter expert peer tutoring remains the gold standard for developing the student's 'zone of proximal development', the stretching of what they can learn on their own to what they can learn with assistance (Vygotsky 1978), using non-subject matter expert tutors may not be an adequate response. Instead, clearly defining the role of the tutor and his/her function as part of the collaborative learning framework from the onset may mitigate to a certain extent some of the potential impediments that tutor behaviour can pose to the collaborative learning process. When a tutor is recruited by InZone (primarily on the recommendation of the course lecturer who knows the tutor through their own classes at the University of Geneva), they receive the earlier mentioned training on how to deliver online tutoring for refugees. This training includes clear explanations of the roles and responsibilities of a tutor in the learning ecosystem. The roles and responsibilities of each actor in the learning ecosystem are clearly defined in advance and shared among the constituents of the learning ecosystem. This effort helps to ensure that each person acting in the ecosystem knows what they need to do and what their colleagues do, so that the ecosystem can function correctly.
Despite the efforts made to mitigate for deviations in individual tutor behaviour, online tutoring can be greatly affected by subtle idiosyncrasies shaped by the cross-cultural factors. In order to address them, InZone's tutoring programme first confronts them head on in the training of new tutors in an effort to give them the tools they need to recognise these factors and then mitigate their impact on their tutoring.

Cross-cultural factors
Taking Geert Hofstede's 1980 Cultural Dimensions theory as a theoretical framework for understanding national culture, we can gain an insight into how communication across cultures can be influenced by general idiosyncrasies typical to a specific culture and thus in turn influence the student-tutor relationship. Hofstede's theory poses five separate dimensions (individualism-collectivism, power distance, masculinity/femininity, uncertainty avoidance, long-term orientation and indulgence), which act as a broad rule of thumb for quantifying typical behaviour in a culture (Hofstede et al. 2010). This theory forms the theoretical 'backbone' of InZone's training approach to understanding cross-cultural behaviour and how tutors and students can best navigate the cross-cultural factors that impinge on their pedagogical interactions. The subsequent sections give a brief explanation of each dimension illustrated with an explanation of how each may affect the student-tutor relationship. 5 Individualism versus collectivism. Individualistic cultures tend to promote the 'I' over the 'We' and tend to hold a strong preference for self-reliance. Collectivistic cultures pay more attention to the 'We', and membership with a social group is more highly prized. This can manifest in the student-tutor relationship when students from a collectivistic culture are reluctant to speak up and take part in discussions. Conversely, students from more individualistic cultures come from academic cultures where active oral participation is the norm and thus collaborative learning may be facilitated more easily.
Power distance. Power distance symbolises how we relate to others in perceived authority. Students from cultures with high power distance tend towards a formal relationship with tutors (as they perceive tutors to be authority figures), which can manifest with a certain amount of rigidity in terms of addressing the tutor and a reliance on the tutor as an expert. Students from low power distance cultures may be more inclined to view the tutor as a peer and thus feel more comfortable addressing the tutor on a first name basis.

(page number not for citation purpose)
Masculinity versus femininity. Masculinity refers to cultures which praise typically 'masculine' qualities such as competitiveness, achievement and success over what are perceived to be more 'feminine' attributes such as cooperation and being caring. In a student-tutor relationship, a student from a more 'feminine' culture may respond better to more inclusiveness and encouragement rather than tasks that are designed to highlight individual achievement.
Uncertainty avoidance. This dimension relates to how we respond to the level and rigidity of rules in society. In low uncertainty avoidance cultures, rules exist, but their strict observance is not mandatory. In high uncertainty avoidance cultures, rules are made to be followed. Students from high uncertainty avoidance cultures may require more instructions from their tutors and greater structure in delineating the tutoring relationship than students from low uncertainty avoidance cultures.
Orientation to time (monochronic vs. polychronic). In polychronic cultures, time is more flexible than it is in monochronic cultures. In monochronic cultures, tutorials tend to be arranged for specific times and last for a prescribed amount of time. Students from polychronic cultures may face some difficulties in adjusting to this rigidity.
Hofstede's cultural dimension theory has been criticised for reducing culture to an overly simplistic conceptualisation and its broad categorisation of cultures into neat packages (Kirkman et al. 2006). Despite the criticisms, it remains the principal theory guiding intercultural communication and can provide a basic framework, which may be applied to the student-tutor relationship to better understand some of its dynamics. Tutoring programmes that take the time to understand the different cultural idiosyncrasies that come into play in cross-cultural communication may stand a better chance of mitigating potential intercultural pitfalls and achieving the goal of enhancing collaborative learning. InZone employed this theory when developing its training for online tutors as a starting point from which tutors can begin to understand their own behaviour and that of their students. While it serves as a useful framework in a basic sense, the ongoing tutor support that is given throughout the lifecycle of a course from the course coordinator adds an additional avenue for understanding behaviour in an attempt to mitigate shortcomings and optimise interactions.

Online communication
Online communication is recognised as differing significantly from in-person communication. Extensive research has been carried out over the last few decades exploring the differences and their effect on interpersonal communication. The general consensus is that the absence of non-verbal cues, lack of warmth and less demand for engagement in online communication can result in impersonality, shallow interactions and extra difficulty in establishing strong interpersonal relationships (Lee et al. 2011). When coupled with the idiosyncrasies of cross-cultural communication, where non-verbal cues are less readily available, online communication may potentially lead to communication problems that online tutors need to be aware of.
The medium (or media) of communication used in online tutoring (in InZone's case a WhatsApp messaging forum) may have an impact on the formation of the relationship between the online tutor and the student and the relaying of information between the two parties. Media richness theory (Daft and Lengel 1986)  Similarly, social presence theory (Short et al. 1976) suggests that the greater amount or richness of media used allows for greater warmth between the communicating parties (Sherman et al. 2013). Thus, considering these theoretical assumptions, the greater variety of sensory facilities employed in online tutoring can lead to stronger relationship formation and better relaying of information between parties. This in turn may affect the collaborative learning experience. Combining online video communication, written forums, voice messaging and so on (as available in WhatsApp and employed in InZone's tutoring programmes) may mitigate the lack of cues that online communication operates at in comparison with face-to-face communication.
A final consideration for online communication that warrants a mention in the context of delivering higher education to refugee learners is how accustomed each party is to digital communication. As explained earlier, tutors are generally chosen based on the recommendation of the course leaders who have taught them in similar courses at the University of Geneva or our partner universities in North America. These tutors have grown up in a highly digitalised world with ICT routinely part of their studies as well as social interactions. 'Digital natives' (such as these tutors) are very comfortable with ICT (Eisner 2004) as opposed to non-digital native students who have not grown up with such routine. The refugee students who have enrolled in InZone courses do not have the same customisation to ICT in their daily lives. Although access to a computer and internet is a prerequisite to take part in InZone courses, access to either is a rare and expensive commodity in a refugee camp. InZone has thus installed Learning Hubs in locations that it works in (reconditioned shipping container with computers and access to the internet) and provides students with basic instructions on how to utilise them (Carron 2019). This mismatch between the level of familiarity and access to technology between tutors and students and its potential impact on online communication skills is a key consideration that InZone highlights in the training it offers to new tutors.
Overall, the peculiarities and particularities of online tutoring are determined by various factors including tutor behaviour, cross-cultural communication, method of communication employed, relationship development between the online tutor and the students and level of experience of tutors and students. These factors coalesce to influence the online communication dynamic between the student and the tutor and can ultimately play into the collaborative learning experience. To better understand how the role of online tutors support learning for InZone's refugee students, the subsequent section presents data from nine InZone courses held at Kakuma and Azraq refugee camps. Table 1 illustrates data collected from nine InZone courses that took place between 2017 and 2018 in Kakuma and Azraq refugee camps. The data were collected and collated on-site in both refugee camps by InZone staff over the 2017 and 2018 academic years. Due to ethical considerations for refugee learners and the difficulty of gaining access to the camps, it was not possible to collect detailed individual information relating to personalised pedagogical assessments. However, these data represent a broad overview of who studied which courses with InZone and how their learning was supported. For the purposes of this case study, these data are sufficient to illustrate the number of students and online tutors employed to support their InZone programme experience.

11
(page number not for citation purpose)   In all courses, online tutors (n = 21) played the chief pedagogical role, interacting online with the students (n = 118) at least one time per week for 2 h at a time over WhatsApp synchronously. The time spent on asynchronous interaction over WhatsApp was not monitored.
Although the number of online tutors employed in each course and the extent of their pedagogical involvement differed, the essence of each online tutor's role in their courses was consistent with the general role description of the online tutor outlined earlier in this article. Averaging the course completion rates for all nine courses, the figure obtained is 82.56%. In a synthesis of literature looking at online retention rates, Baiwa (2016) points out that between 40% and 80% of online students drop out of online courses. 'InZone courses' higher completion rate to the majority of other online courses may be due to a variety of reasons (e.g. choosing the right students for the right courses), but when we look at the main reasons for high dropout rates, as pointed out by Baiwa -maintaining motivation, technological capacity and social issues -it is a fair assumption that the tutors played a positive role.
The dropout figure for MOOCs may be even higher (90%) than Baiwa's predicted 40%-80%, according to a report commissioned by the Erasmus programme in 2016 (Mook-Maker 2016). While no one consistent figure represents all blended learning approaches (by their very nature, they constitute a 'recipe' of a certain percentage of face-to-face and a certain percentage of online pedagogy depending on each particular course's individual make up), it appears that retention rates for blended courses in general appears to be higher than online courses. Authors such as Bonk and Graham (2012) point out that the increased interaction and motivation afforded by blended approaches may help to augment retention. Bonk and Graham (2012) further suggest that proactive blended approaches to meet the needs of students, rather than reactive approaches that try to 'fix' the problem of dropout rates, is effective in retaining students. This is consistent with InZone's approach to online tutoring and the training the tutors receive. The main role of the InZone tutor is one of a pedagogue and as such, is partially a pastoral role that helps to maintain motivation through understanding the students' particular social settings and technical capacities.

Conclusion and recommendations
This study set out to explore the role of online tutoring in enabling higher education in refugee contexts and to discuss some of key factors that may affect its functioning in a collaborative learning model. By laying out the role of the online tutor in the context of InZone's collaborative learning ecosystem model, three main factors that affect online tutoring and may in turn affect collaborative learning were identified: tutor behaviour, cross-cultural awareness and online communication. The particularities and idiosyncrasies at play in these three factors play a role in moulding the student-tutor relationship. InZone has attempted to address the potential positive and negative impacts of these factors through a dedicated training programme for online tutors, which is tailored to the needs of the refugee context in which it works. Such a feat is by no means an easy task and may be more possible in programmes with smaller student numbers (InZone class size in the course data was 13). However, the lessons that can be learned from the experience of the InZone tutoring programme can serve as inspiration for other higher education providers who work in similar settings. For instance, motivation, a key factor in lowering student attrition, 13 (page number not for citation purpose) can be facilitated by tutors who understand their students' situations and have a good rapport with their students.
As the development of online tutoring programmes for delivering higher education to refugees in fragile contexts is a relatively new area of study, little research and analysis has been carried out on its usefulness for refugee learners and its long-term impact on their education attainment. While carrying out research with refugee learners is not always a straightforward task (for example obtaining ethical clearance to do so can take up to a year -depending on the camp administration and the country in which one operates), closer observation of refugee education contexts and deeper research into the impact of online tutoring for refugees have the potential to yield important findings for the future of pedagogy in such contexts. In addition, developing this field of study has the potential to bring a wide range of rewards to universities in the west and their students who engage in such programmes. To understand how best online tutors can improve learning in refugee contexts, it is recommended that more in-depth and personalised research be carried out to assess the needs of individual learners and their communities.