Aims The research reported here is premised on the belief that lying at the heart of the process of education is the interpersonal relationship between the teacher and learner. However, we know very little about the teacher-pupil relationship from the view of ethnicity, particularly from the pupils' perspective. It has been argued elsewhere (Naylor, 1996) that the revelation of these perceptions and conceptions is important in understanding the effects teacher racism may have on pupils, why some schools alienate so many ethnic-minority pupils, and why educational opportunity is not equal for all ethnic groups. The exploratory research reported here, which is part of a larger project (Naylor, 1996), has attempted to investigate adolescent pupils' conceptions of teachers' racist behaviour and actions in the classroom.
Methods Two interrelated methods of data collection were used.
Method I A five-minute video presentation of an actual classroom episode, reconstructed for the purposes of the research, was produced. The film was scripted and role-played by a multi-ethnic group of twenty-one Year 9 and 10 (13-15 year old) English comprehensive school pupils. The film focuses on the behaviour of a white female teacher towards pupils of `visible ethnic minority group membership' (VEMGM; Cohen and Manion, 1994, p. 268). A drama teacher elicited performances from the pupils and faithfully interpreted their directions in her portrayal of the devised teacher role. In order to elicit the role-players' perceptions of what the film was intended to portray, and also to provide a check on the internal validity of the film, the role-players subsequently provided their independent accounts of the roleplay.
Method 2 An opportunity sample of sixty-one mid-adolescents (Year 9) responded to the role-play, using a bubble dialogue `comic strip' technique. This technique involved 'grabbing' eight or nine key still images from the video film. 'Say' bubbles of the actual speech from each shot were superimposed on to the still images. Empty 'think' bubbles were also provided for the respondents to complete, thereby permitting them to participate in the private worlds of the characters in the role-play (O'Neill and McMahon, 1990). The qualitative data derived by this method were subjected to a four-step qualitative content analysis (Strauss and Corbin, 1990, pp. 63-5; Glaser and Strauss, 1967) and to systematic network analysis (Bliss et al., 1983). These data have been used to model diagrammatically the respondents' ideas about how they think the racist teacher thinks.
Outcomes Method I The adolescent role-players subsequently named the role-play of classroom teacher-pupil interaction 'Borrowing', since the plot centres on one pupil borrowing another's eraser. …