To Answer or Not to Answer: Predicting Item Nonresponse in School Self-Evaluation

• In Event: Marriott Roundtable Session 7
  In Roundtable Session: Applied Methods and Measurement in Education Research

Abstract

Data of school self-evaluation (SSE) are often used to inform decisions or actions in light of quality and equity in schools. These data on a school’s processes are often collected by administering questionnaires among teaching staff. When respondents drop out for one or more items, referred to as item nonresponse, it may lead to distorted results. This study aims to explore to what extent respondents’ characteristics (motivation to complete the questionnaire and tendency towards socially desirable responding), item characteristics (design) and evaluation perspective (accountability vs. development) can predict item nonresponse. A quasi-experimental SSE was set up with 378 participants. Cross-classified multilevel analyses show that item design and evaluation perspective indeed predict item nonresponse. Implications for the SSE field are discussed.

Authors

• Jerich Faddar, University of Antwerp

• Jan Vanhoof, Antwerp University

• Sven De Maeyer, University of Antwerp