This study involved close observation of 26 pupils aged 9-14 years in four schools over a period of 12 weeks, and then a further study involving 111 secondary school pupils, carried out over three years. It investigated the influence of prior knowledge on children's understanding of ideas in science, in three related areas: evidence for and against the contention that pupils formulate 'alternative frameworks' of interpretation; the pupils' understanding of the purpose of their work; and the influence of context upon meaning. Findings include (1) a caution against the rigid postulation of alternative frameworks, (2) indications, nevertheless, that pupils do anticipate and explain observations by constructing tentative predictions from prior experience, and (3) evidence that pupils often invent their own purposeful aims for lessons. During the investigation several issues relating to the teacher as researcher emerged, and the thesis chiefly reports this over-riding aspect of the inquiry. The study highlights and appraises implications for teacher-researchers in the following areas: the theory base of the research; the choice of questions to be investigated; the choice of methods; means of improving reliability and validity; and the practical opportunities for teachers. The thesis suggests ways for teachers to guard against such dangers as eclecticism, subjectivity and lack of perception. It compares the strengths and weaknesses of qualitative methods, includes an interview checklist, and compares the contributions which teachers and non-teachers can make to research. The study also suggests the need to develop semi-structured interviews in classroom research, and argues for a combination of the expertise of teacher- researchers and professional research teams in tackling inquiry into concept development in science.