The desire for accountability has led designers of large-scale writing
assessment to embrace a formal semantics theory of writing evaluation
that disregards cultural influences on the writing and marking proces
ses in its emphasis on consistency and unbiased scoring of student wri
ting. However, an examination of the scores assigned to the narrative
writing of elementary and middle-grade students on large-scale examina
tions in Canada, Great Britain, and the United States reveals the need
to consider sociocultural influences because scores on girls' writing
are consistently higher than scores on boys' writing. This study inve
stigates the relationship between teachers' perceptions of gender-rela
ted differences in grade six students' narrative writing and teachers'
scoring of five narrative papers written by sixth-grade boys and girl
s. Participating teachers scored the narrative papers using a Canadian
provincial scoring guide. The teachers who did not know the gender of
the writers were asked to identify the gender of the writer (if possi
ble) and to describe the characteristics of the narrative that helped
them to determine the writer's gender. All teachers were asked to comp
are and contrast girls' and boys' classroom narrative writing. Signifi
cant differences in the scores appeared for only one paper, written by
a girl, that exemplified both male and female narrative writing chara
cteristics. Teachers who felt that a girl wrote this paper scored the
paper significantly higher than teachers who identified the writer as
a boy. Teachers who disagreed in their identification of the writer's
gender drew upon similar elements from the writing to support their vi
ews, yet evaluated those elements in contrasting ways that revealed a
stance privileging girls' narrative writing. In addition, teachers cha
racterized girls' classroom writing as being more sophisticated than b
oys' writing on all five dimensions of Moffett's (1968) continuum of d
iscursive growth.