The influence of contrast effects upon teachers’ marks

Abstract

Summary A marking exercise is described in which 336 teachers evaluated three samples of written work, which differed in quality and authorship. The sequence of the work samples was rotated such that each piece of work was seen and marked in each of the three possible positions. This allowed comparisons to be made between the marks awarded to work of varying standards when it preceded or followed other work of contrasting quality. There was a tendency for the good piece of work to be assessed more favourably when it followed work of a lower standard than when it preceded such work. Conversely, the poor quality work was assessed more harshly when it followed, rather than preceded, work of contrasting quality. Furthermore, two samples of work preceding work of a contrasting quality produced greater biasing effects than a single sample.