when setted (means: female 2.21, male 2.33; t = 2.1, df = 1,1463, p = .03). Males
agreed more strongly that setting leads teachers to ignoring the range of abilities in the
class (means: females 3.4, males 3.15; t = -4.3, df = 1, 1540, p = .0001) and that only
very good teachers can teach mixed ability classes successfully (means: males 2.85,
females 3.0; t = -2.6, df = 1, 1534, p = .01 )
Using the attitude to setting measure as the dependent variable, step wise multiple
regression was undertaken to establish which factors would best predict teachers’
attitudes. Two factors were found to be important. The first was type of school, mixed
ability, partially set or set, (standardised beta weight .225) the second was the subject
taught (standardised beta weight .078). This gave a multiple R of .239 accounting for
slightly more than 5% of the variance (F = 29.13, df = 2,962, p = .0001).
When multiple regression was undertaken for type of school separately, slightly
different patterns emerged. In the Set schools the significant predictors were the
individual school (beta weight -.128) and the subject taught (beta weight .109) (R =
.167, F = %.14, df = 1,359, p = .006). In the partially set schools length of time
teaching at the school was the only predictor ( beta weight .175) (R = .175, F = 9.36,
df = 1,298, p = .002). In the mixed ability schools the subject taught ( beta weight
.21), the length of time teaching at the school ( beta weight -.126) and whether the
teacher had a higher degree ( beta weight -.119) predicted attitude toward setting (R =
.274, F = 8.08, df = 1,300, p = .00001) The shorter time the teacher had been teaching
in a mixed ability school the more positive their attitudes towards setting. In relation
to a higher level degree, e.g. masters or doctoral qualification, the higher the level of
16