Unreliability of Marking: further evidence and a possible explanation

Abstract

ABSTRACT Marking of essays and examination papers is known to be unreliable in several ways. Some further evidence of this lack of agreement by markers is presented from an investigation of naturally occurring student ‘malpractices’. In a search for explanation of the causes for marker unreliability, a second study investigated the effects of examiner personality. It was found that the marks given to an essay were unrelated to E, N or P scores on the Eysenck Personality Questionnaire but were positively correlated with L scores. This finding raises interesting hypotheses about the effect of staff‐student interactions on biases in marking. In particular ‘social desirability’ seems to be a variable worth further investigation.