Equally assessed? Influence of home background and gender on teacher assessment of student reading proficiency in relation to level of achievement in PIRLS 2001

• Abstract It is well known that gender and social background play an important role for a students’ performance in school. Previous research has suggested that girls outperform boys in reading and writing and that students from higher SES perform better than students from families with lower SES, which often implies lower economic and cultural capital. Thus, the outcome is strongly correlated with gender and home background, but what about the teacher assessment? The aim of this paper is to investigate the teacher ratings of students’ literacy skills in relation to the PIRLS 2001 test scores. One purpose is to determine if the teachers’ ratings follow student development and especially study the ratings in relation to gender and home background. As a natural extension of the PIRLS 2001 study, Swedish teachers rated students on 18 different statements about students’ literacy skills on a scale ranging from 1-10. This enables an investigation of the relationship between teacher ratings and student achievement in reading literacy. The analysis was conducted in three steps. First, a two-level model was fitted to the data and the variance of the teacher ratings were studied by means of a latent model of the teacher ratings. Second, students’ achievements on the PIRLS test were used to examine the variation in teachers’ ratings, both within and between classes. Third, students’ gender and home background were introduced to investigate differences in teachers rating practices. Results show that the variance among the teacher ratings was high both within class and between classes, but that a substantial part of the variance was accounted for by student achievement.

Subject headings

SAMHÄLLSVETENSKAP  -- Utbildningsvetenskap -- Pedagogik (hsv//swe)

SOCIAL SCIENCES  -- Educational Sciences -- Pedagogy (hsv//eng)

Keyword

Teacher assessment

large-scale studies

Publication and Content Type

vet (subject category)

kon (subject category)