Traditionally, supervisors work with students on an individual basis and postgraduate development programmes are run on site. However, with increasing numbers of part‐time and international students, supervisory relationships are likely to be conducted at a distance as students study alongside other commitments. Isolation can often be a key feature for many postgraduates, whether based in the same institution as the supervisor and more particularly for international students or those studying at a distance. It can also be an issue for their supervisors. However, in this age of electronic communication, interactions at a distance should be able to be at least as robust as many of those conducted face to face. More broadly, the international research community is something generally cited as the context for research work, supervision, study and publication more generally. Entry firstly into the university academic community, and then into the larger global research community can be enabled, we would argue, by the support of communities of practice from the outset of postgraduate students’ and supervisors’ interactions. In this context, considerable numbers of international postgraduates at Anglia Ruskin University (UK) and their supervisors are being effectively supported through three innovations which build institutionally related communities of practice: (1) Guardian supervisors work with the students on research development programmes, with accompanying meetings organised during and round programme workshops, which focus throughout on strategies of meta‐learning. Subsequently they support students’ work at a distance with emails, and webcam correspondence. (2) PhD students are empowered to develop mutual, critically focused support for each other’s work through the enhanced use of the cohort in the compulsory research development workshops and through ongoing discussion lists, self‐help groups and symposia. (3) Supervisors of international postgraduates are supported as a community of practice through the revision of online supervisory discussion and development. This paper is based on action research carried out with current and graduate students, guardian supervisors and supervisors as collaborators to explore the rationale, problems, practices and the richness of the experience of working with a system which fosters communities of practice, involving guardian supervisors, distance supervisors and postgraduate cohorts.

Results from action research conducted at APU largely with Israeli PhD students, start to identify good practice in supervisory dialogues, peer support and in facilitative strategies as part of research development programmes. Our results, focusing on learning conversations between students and supervisors, students and peers on research development programmes, identify strategies, supervisory relationships and support, which aid completion. This research builds on earlier studies (Wisker, 1999; Wisker et al ., 2003) into PhD students' development of research methods and conceptual frameworks, and on the work of others (Moses, 1984; Pearson, 2000) into supervisory relationships. Focus groups, workshops, questionnaires and individual supervisory dialogues with PhD students, have been conducted using an action research format. This paper concentrates on elements of the action research taking place at completion stage in students' PhDs and during the research development workshop programme (stage 3, 2000, 2001, 2002)

Most research indicating dissonant forms of student learning engagement, leading to problems in the achievement of learning outcomes, is with undergraduates. Action research at Anglia Polytechnic University (APU) involving questionnaires, focus groups and supervisory dialogues, conducted with Israeli and UK postgraduate students between 1998 and 2001, indicates that dissonance in research seen as a form of learning produces potentially significant difficulties for students at different stages in their work. These difficulties emerge principally when developing a proposal, deciding on research methodologies and methods, undertaking the research, and developing and maintaining links between findings, analysis and conclusions, specifically those which aim to lead to transformative research outcomes. A research development programme and supervisory dialogues provide support, enabling some PhD students to overcome dissonance and succeed in their studies.

Much research into doctoral student-supervisor relations focuses on developing positive interactions. For many students, however, the research experience can be troubled by breakdowns in communication and even the loss of the supervisor(s), turning the student into a doctoral ‘orphan’ and impacting on their academic identity and ability and confidence in producing a sound doctoral-level contribution to knowledge. Our work with a range of UK- and internationally-based doctoral students looks specifically at reasons for supervisor loss and/or absence and the students' experience of being doctoral ‘orphans’ in terms of identity, confidence and progress. In focusing on those who achieve successful completion, it suggests the need for institutional and community support and highlights the development of effective strategies leading to ownership, empowerment and emotional resilience.