Abstract
Surveys are frequently used to inform consequential decisions about teachers, policies, and programs. Consequently, it is important to understand the validity of these instruments. This study assesses the validity of measures of instruction captured by an annual survey by comparing survey data with those of a validated daily log. The two instruments produced similar rankings of the frequency with which teachers use particular practices but more than three fourths of the teachers in the study were found to overreport their instruction on the annual survey. Multilevel models revealed a number of teacher and school characteristics related to survey reporting error. The study’s implications for users of survey evidence are discussed.
Get full access to this article
View all access options for this article.
