Abstract
The sociology of education has an ambivalent relationship with IQ and 'ability' tests. Some accounts of educational inequality incorporate measures of intelligence, others restrict their explanations to the 'sociological' rather than the 'psychological' variance, and still others reject psychometric theory in its entirety subjecting it to a vigorous critique. There are evident contradictions in this area. Studies that show environmental (rather than genetic) effects on 'ability' test scores and school attainment are welcomed, while 'ability' is nevertheless regarded as a 'social construct'. Standardised tests produced within the psychometric paradigm are employed in certain research applications even as their theoretical rationale is dismissed. It is argued that this situation is unsatisfactory and that a realist account of cognitive socialisation based on the work of Bernstein and Vygotsky would represent an advance in our capacity to explain social differences in educational attainment.