The Learning in Science Project (1979-1981) studied the problems and difficulties of learning science at the 10-14 year old level in New Zealand to provide insights into the world of the science learner and provide an alternative perspective of the science classroom to that currently held by teachers and others. This paper offers practical suggestions on initiating and carrying out classroom interview data collection procedures used in the Learning in Science Project. The suggested do's and dont's of structured interviews, based on experiences in the Project discuss: (1) establishing contact with the teacher, (2) establishing contact with the pupil and (3) interviewing with the Interview-about-Instances method--a procedure for establishing a person's understanding of a particular word or the concept(s) a person associates with a particular word. (PN)