Who responds to surveys assessing teachers’ specialized knowledge to teach literacy?Implications for generalization and representation

The growing research base concerning the role of teachers’ specialized knowledge to teach literacy in supporting students’ reading achievement—and policies based on this research—primarily relies on data from teacher knowledge surveys. We examined the possibility that teachers who respond to knowledge surveys are systematically different from teachers who only partially or do not respond. In this exploratory study, we examined response patterns (complete, partially complete, no response) to teacher knowledge surveys in three U.S.-based projects and explored potential differences among response patterns based on sociodemographic characteristics, teaching beliefs, and classroom/program characteristics. We identified response patterns using missing value analyses and applied the What Works Clearinghouse guidelines for group equivalency. Collectively, these projects represent data from ~ 870 inservice and preservice preschool-Grade 5 teachers. Across datasets, 84% of teachers completed the surveys, 10% partially completed, and 7% were nonresponders. We identified meaningful nonequivalence across response patterns for the majority of characteristics examined. Our findings suggest that teachers who complete knowledge survey measures meaningfully differ from those who partially or do not complete them, affecting the validity and generalization of conclusions. This also underscores the need for researchers to use best practices in survey design as well as modern statistical methods for handling missing data and accounting for nonequivalence. We hope these findings lead to greater consideration of representation in future teacher knowledge studies.