Teachers, she says, do not appear to distinguish between pupils’ academic abilities and levels of interest. “This seeming failure to recognise pupil behaviours that reflect levels of interest suggests that the teachers are not teaching towards interest goals.

“If you wish to identify the educational factors that influence pupils’

attitudes to science, we must look in detail at those factors that have an impact on the pupil inside the classroom, rather than at the general and organisational features of schools and classes which, though intuitively attractive, have been shown (in her study) to be anything but fruitful.”

Variables that might help to explain differences in pupils’ attitudes to science and tell why pupils’ feelings about science become so negative by the end of S2 are likely, Dr Brown says, to be provided by:

* An examination of the nature of teacher-pupil interaction in science classrooms.

* The teacher’s patterns of communication with individual pupils and groups of pupils.

* The transmission of the teachers’ expectations to the pupils.

* The particular topics that are covered in the lessons.

* The strategies and tactics within strategies adopted by the teacher.

Dr Brown says her study has provided evidence to suggest that in Scotland no substantial influence on the development of pupils’ attitudes to science is exerted by size of school, whether the school is inner-city or not, the number of periods given to science, the size of class, mixed ability grouping or streaming, mixed or single sex classes, integrated science policy or separate subject teaching.

“The most obvious candidate is teacher influence . . . Although the relationship between teachers’ attitudes and pupils’ attitudes, if it exists at all, is elusive, there is some evidence that teachers’

expectations of their pupils’ academic performance has an influence on attitude development (particularly interest in science).”