This paper reports on a study of 100 newly qualified secondary teachers in New Zealand as they constructed and explained images of their ‘best teaching’ at three points over the first eighteen months of full-time employment. Images can be a key to understanding how teachers see themselves, whether they see themselves performing a role or whether they have embedded an identity as teacher. Metaphor/simile/image elicitation targets three components of personal interpretive frameworks outlined by Kelchtermans (1993). These are the descriptive self-image, the evaluative self-esteem and the normative task-perception. We argue that through image expression and subsequent reflection on these images, teachers can better understand and interpret their teaching practice (Haigh & Ward, 2004). It is also possible that in the elaboration of the image the teacher also reveals both knowledge and belief aspects of his/her subjective educational theory (Elbaz, 1981; Kelchtermans, 1993). Images may also reveal a teacher’s philosophy of teaching (Bérci, 2007). Image production through elicitation may also encourage identity construction (both consciously and subconsciously) through increased teacher reflexivity around role and identity as they engage in negotiation of meaning through “semiotic activity in communities of practise" (Mantero, 2007, p. 377). Since a teacher’s identity is impacted by situational factors and negotiated through language, encouraging an elicited image may help teachers to understand themselves as teachers in their particular context and “can stretch their imaginations to consider alternative conceptions of teaching and self-as-teacher” (Bullough & Stokes, 1994, p. 220). The importance of image expression has long been recognised internationally within teaching and teacher education where elicited image construction and/or analysis of the images used in reflective story telling has been a consistent part of educational teaching and research across the globe over several decades. As well as encouraging practitioners to reflect on their own metaphors, Munby and Russell (1990) also argued that researchers can “explore practitioners’ metaphors to gain insights into how they [construct] their professional worlds” (p. 117). Studies that encompass/accept forms of image expression in addition to pure metaphor or simile have the potential to contribute to our understanding of teachers’ thought processes, development of the professional self and teacher identity formation. Our review of the international literature identified few reports of the use of metaphor, either from narrative analysis or elicitation, over extended periods. In contrast, the study reported on in this paper is a longitudinal study that offered the opportunity to trace changes in the beginning teachers’ imaging of themselves as teachers. The questions guiding the analysis of the data for this paper were: • What images of “best teaching” do these newly qualified teachers present? • How do these images change with experience? • What insight do these images provide into the teachers’ developing/evolving professional self or identity?

The findings reported in this paper are drawn from a longitudinal national project Making a difference: The role of initial teacher education and induction on the preparation of New Zealand secondary teachers (Anthony et al., 2008). Semi-structured interviews were conducted with a cohort of 100 teachers at the mid- and end-point of the teachers’ first year of teaching and midway through their second year. Each interview included a question designed to elicit the teachers’ images of themselves when they were “teaching at their best” (Palmer, 1998, p. 148). They were prompted for a simile or metaphor but some responses were more generally descriptive. The proffered images from each interview were treated as separate data sets and then subsequently compared to illuminate changes. Imagery categories were identified. Where ambiguities arose the elaborated answers were consulted to check the participants’ meaning. Reliability of categorisation was strengthened through having two researchers categorise independently.

76 of the 100 NQTs interviewed provided a complete set of images across all three interviews. Figurative images of ‘best teacher’ were grouped into five ‘root image’ categories. A few images appeared to be able to be categorised within more than one of the root categories. If this was so the elaborated explanations were analysed carefully and the best fit found. Only 20 of the 76 teachers did not change the ‘root image’ in the representation they presented across the three interviews. The most frequently proffered images at Interview #1 were those that evoked an action or a person experiencing and expressing an emotion. The next most frequent were images of an expert guiding the learner through new territory and the teacher as a performer. A small number (n=5) were categorised as power sharing, where the evoked image was one of careful and supportive oversight of the student’s learning from a distance. These image categories were then analysed for explanations of the teachers’ philosophies of teaching and learning. A more detailed description of the teachers who presented a stable or evolving view as power sharers was developed with the view to understanding factors contributing to the teachers’ images of self.