Teachers' written feedback comments on narrative texts in Elementary and Secondary Education.

Highlights • A classification related to teachers' responses to students' texts is elaborated. • Profiles of assessment practices in Elementary and Secondary Education are proposed. • Assessment related to spelling and grammar is the best predictor of global grades. • There are relevant differences in assessment profiles in both educational levels. Abstract One of the most widespread practices in compulsory education is the annotation of feedback messages by teachers on texts written by students. The aim of this paper is to describe the nature of these types of teachers' responses, in order establish possible profiles of assessment practices in Elementary and Secondary Education, as well as to identify which aspects of writing a story are considered more important in writing assessment. An analysis was made of 10,585 assessment messages which 41 schoolteachers in Elementary and Compulsory Secondary Education wrote on short stories composed by 393 students from various schools. The assessment notes were categorized according to code, place, extension, assessment content, and implicit metalinguistic content. Findings reveal the predominance of direct correction of specific spelling and grammar mistakes. The frequency of these corrections is, additionally, the best predictor of the global grade given by the teacher for the composition. Therefore, we conclude that teachers of compulsory education approach assessment of narrative texts from a conception which places excessive emphasis on more local and superficial aspects of the composition in detriment of other semantic, rhetorical or pragmatic aspects. Some teachers, however, mostly in Secondary Education, also recorded non-corrective assessment content, such as markings, questions, suggestions for expansion, or justifications. Certain assessment patterns are evident, which combine other types of evaluation (semantic-organizational, or superficial). Finally, the limitations and implications of those results toward teacher training are discussed. [ABSTRACT FROM AUTHOR]