Sex Bias in Student Assessment Overlooked?

Newstead & Dennis (1990) attempted to replicate a study by Bradley (1984) which demonstrated sex bias by second markers of university undergraduate students’ projects relative to marks awarded by supervisors. Newstead & Dennis did not find the hypothesised differences between supervisors’ and second markers’ marks in their study and took this to indicate that neither marker demonstrated sex bias in their polytechnic department. However, they overlooked the evidence in the data they presented which supported the view that both markers might be equally biased. As would be expected if markers were equally biased, there were fewer marking disagreements across class boundaries than had been found in Bradley's study. Those disagreements that were found by Newstead & Dennis were more likely to be resolved upwards for male candidates than for female. Data from a second department where second markers marked anonymous scripts were included but Newstead & Dennis did not compare the standard deviations of female students’ marks in the two departments and these differed significantly in a manner indicative of sex bias in Newstead & Dennis's department.