Appraising Validity Log Data
Validity is an important issue for appraising the significance of evidence on instruction, such as that offered in this symposium. Many factors can interfere with the quality of data on instruction, from the imprecise and variable meaning of the language used to describe teaching, content, and instructional goals, to the tendency to over- or under-report particular sorts of practice (e.g., instruction focused on problem-solving). This paper compares firsthand observations of mathematics teaching with log data collected on the same lessons by observers and teachers. We appraise match rates between two independent raters of each lesson, and between those raters and the teacher. Results suggest that validity of these log data used in this study were more uniform on some dimensions of mathematics instruction than others. For example, evidence on more complex mathematical goals and instructional practices were more likely to vary as a function of knowledge and interpretation of terms, whereas evidence on content covered, time spent, representations used, and student activity were more likely to be consistent across data records.
Keisha Ferguson (United States)
Graduate Student Researcher
The Study of Instructional Improvement
The University of Michigan
Keisha Ferguson is a fourth year graduate student at the University of Michigan studying mathematics and teacher education. She currently serves as a graduate researcher on the Study of Instructional Improvement collecting data in first grade classrooms as well as conducting a mathematics validation study. For over ten years, Keisha served as an elementary school teacher, math specialist and teacher consultant.
(30 min. Conference Paper, English)