PROCEEDINGS OF THE 21ST ANNUAL CONFERENCE ON RESEARCH IN UNDERGRADUATE MATHEMATICS EDUCATION
2018
San Diego, California
First Results From a Validation Study of TAMI: Toolkit for Assessing Mathematics Instruction
Page: 727
Many researchers consider observation to be a ‘gold standard’ for measuring classroom practices since self-report surveys may be prone to bias. In this paper, we explore how the design of survey instruments and observation protocols affects the trustworthiness of the data collected. We describe our process of developing well-aligned observation and survey instruments in order to reduce sources of measurement error. We present results from a large-scale test of these instruments in 176 observations of 17 different math courses. Our results indicate that when survey instruments are designed to describe what happens in a course, rather than evaluate the quality of the instruction, and when those survey results are compared to observation protocols measuring teaching in the same way, self-report surveys are largely trustworthy.