The main route into the teaching professions is some kind of adequate formal training. However, in recent years registration of teachers has become an additional tool for policymakers to make a clearer distinction between qualified and non qualified teachers. The process of becoming fully registered as a teacher is then often connected to a probationary year and induction programs, as is the case for instance in England, Wales, Scotland and New Zealand (cf. Cameron, 2007; Killeavy, 2006). In other countries, as the Nordic countries Denmark, Finland, Norway and Sweden, the issue of who is qualified to teach is regulated by legislation connected to adequate teacher training (Bjerkholt & Hedegaard, 2008). In Sweden, however, a government inquiry was initiated in May 2008, Teachers' Qualifications and Authorisation, Registration and stricter qualifying rules, proposing registration for teachers and a mandatory induction program (Swedish Government Official Reports 2008:52). The proposals of the inquiry are in marked contrast to certain traditional Swedish educational values, e.g. that mentors should not be involved in any formal evaluation of new teachers’ competencies, but that their role should be to promote their professional development (Fransson, & Gustafsson, 2008). The relationship between mentors and mentees is often considered a vulnerable one, based on confidence and trust (McNally & Oberski, 2003). Research also shows that new teachers may avoid questioning colleagues in order not to risk losing prestige or being regarded as incompetent (Rust, 1994). In the inquiry, principals are proposed to be responsible for the evaluation of new teachers’ competencies, and mentors are proposed to participate. This matter has been discussed in some of the 107 responses to the inquiry that were submitted to the Swedish Ministry of Education. The aim of this article is to analyse to what extent the role of mentors and principals in the evaluation of new teachers is discussed in these responses, and in what ways this matter is discussed. The focus is on what kind of knowledge input the responses could give in the processes of governance and regulation.

The data analysed has been the text of the 107 responses that were submitted to the Ministry of Education. The responses cover approximately 320 A4-pages, with the most extensive one being 5 480 words long, and the shortest one merely 29 words. The responses were grouped into eight categories, covering different types of interest groups; Independent schools (4), Unions (15), Local authorities (18), Higher Education institutions offering teacher education (29), Public authorities (16), Institutions specially relevant for registration (9), Other special interests groups (17). These categories made it possible to analyse responses within specific interest groups in relation to other groups and their focuses. The focus of the analysis was what the responses say about the role of mentors and principals in the evaluation of new teachers and how this matter is discussed. Some key features were e.g. content, meaning making, arguments, and relation to research.

The analysis shows that the role of mentors and principals in the evaluation of new teachers was not discussed to any great extent in the responses. Nine of the 107 conciderations mention in some way the dilemmas for principals in the evaluation process, and nine mention dilemmas for the mentors. Seven of these are found within the category of higher education institutions. However, these responses are often shallow without any deeper contribution of knowledge. Only two responses are to be characterized as knowledge-based contributions with any connection to research: one within the category of higher education institutions and the other as a private initiative from a researcher. One conclusion is that questions must be raised in what way more extensive research-based knowledge is expected in these kinds of responses, especially with regards to how rather complex pedagogical questions connected to evaluation, assessment and research-based knowledge are used in policy processes