1 Matching Annotations
- Apr 2016
White (1984, cited by Vaughan, 1991) reported on a study conducted at California State University in which two essays were tucked into a huge sample of essays and read a year apart by the same readers using a 6-point scale. The reading a year later produced scores that were identical to the first in only 20 per cent of the cases. The scores differed by one point or less in 58 per cent of cases and 2 points or less in 83 per cent of the cases. As White points out, a 1-point difference is generally considered unproblematic, but on a 6-point scale the difference between a 3 and a 4 is the difference between a pass and a fail. Obviously, then, changes in examiner severity/leniency over-time have implications for maintaining standards, and must be monitored. Research has been conducted into variations in examiner severity/leniency during the marking of a particular allocation of scripts, a marking period, and over more extended periods of time.
intrarater reliability is only 20%