No abstract

No abstract

This experimental study examined the influence of vocabulary and spelling on teachers' perceptions of English as a second language (ESL) essays. Sixtynine pre-service teachers from Switzerland and Germany rated four upperintermediate ESL argumentative essays and provided written comments on strengths and weaknesses of the student compositions. In Study 1 (N = 34), the level of lexical sophistication and diversity of the four texts had been experimentally varied. Results suggest a successful manipulation, since comments responding to texts with a low quality of vocabulary included more negative feedback points regarding vocabulary. Yet, participants also commented more negatively on grammar when responding to texts with a low quality of vocabulary, indicating a halo effect. In Study 2 (N = 35), the quality of spelling was manipulated in the same four student essays. Results indicate that comments responding to texts with a low quality of spelling involved more negative feedback points regarding spelling (manipulation check) on the one hand, and vocabulary, grammar, and others (halo effects) on the other hand. In both studies, teachers' perception of strengths and weaknesses correlated moderately with their holistic and analytic assessment of the texts. The implications of this study for classroom practice and teacher education are discussed.

When judging subject-specific aspects of students’ texts, teachers should assess various characteristics, e.g., spelling and content, independently of one another since these characteristics are indicators of different skills. Independent judgments enable teachers to adapt their classroom instruction according to students’ skills. It is still unclear how well teachers meet this challenge and which intervention could be helpful to them. In Study 1, N = 51 pre-service teachers assessed four authentic English as a Second Language (ESL) essays with different overall text qualities and different qualities of spelling using holistic and analytic rating scales. Results showed a negative influence of the experimentally manipulated spelling errors on the judgment of almost all textual characteristics. In Study 2, an experimental prompt was used to reduce this judgment error. Participants who were made aware of the judgment error caused by spelling errors formed their judgments in a less biased way, indicating a reduction of bias. The determinants of the observed effects and their practical implications are discussed.

No abstract