Child Respondents: Types and Level of Error Within Questionnaire-Based Judgments of Instructional Quality

• In Event: 32.084 - Poster Session 6
  In Poster Session: 32.084-3 - Learning and Teaching in Social and Cultural Contexts: Methodological Insights

Abstract

Students’ perceptions are often used to investigate instructional quality, also in primary school. Empirical research, however, shows that the assumption that respondents understand and interpret items in the intended way is often not justified (Schwarz, 1999), especially in the case of child respondents (Bowen, 2008). In our study we explored whether items of popular rating instruments are understood and interpreted in the intended way. For this purpose, we conducted cognitive interviews (86 students; third and fourth grades), content analyses, and correlation analyses. Nearly half of the investigated items showed a relatively high level of error. As a conclusion it is necessary to overhaul the popular instruments currently used in practice or research to rate instructional quality in primary school.

Authors

• Gerlinde Lenske, University of Koblenz-Landau

• Anna-Katharina Praetorius, German Institute for International Education Research (DIPF)