Finding Classrooms: Methodology for Identifying Classrooms and Schools for Surveys and Case Studies

• In Event: Present and Future Formative Assessment Practices

Abstract

Formative assessment can be described in many different ways by many different people in education. Test publishers might favor a standardized-based approach, while districts might look at it as an integrated process. In understanding what the current perspective and practice looks like, it’s understandable that many different pictures would emerge in an attempt to describe the current status of formative assessment in mathematics. To gain a cohesive picture of what formative assessment looks like with educators in a classroom setting, it was imperative to strategically select classrooms where we knew some element of formative assessment was taking place. The recruitment and selection methodology we employed for achieving our research goals started by having a firm grasp of our end goal of detailed case study observations in schools with underrepresented students, and walking backward through our other research phases to establish the largest eligible pool of teachers and schools as possible.
The initial research goal of the “Formative Assessment in Mathematics” project was to enroll 120 schools with high percentages of underrepresented populations that are substantially involved with formative assessment policy, instruments, and application for a research and data collection. Initially, we blanketed a basic survey and sent it to all principals in California, hoping that a representative sample would reply and provide us with a representative pool of eligible schools. Unfortunately, survey results came from less than 10% of the schools in California, with many indicating that they did not have knowledge of any formative assessment practices in their math classrooms nor did they fit the student population profile that we were seeking for our sample.
Learning early-on that we were off target, we reexamined our research priorities and created a very specific demographic profile of what we wanted our eligible schools and districts to look like. Through a statistical sorting procedure, we stepped back to the district level and ranked and identified those with desirable characteristics for our study. To identify districts with large numbers of underrepresented students, we used assigned statistical weights to the percent of minorities, percent of free-reduced price lunch students, percent of ELL students, percent of migrant education students, the total number of students in a district, and the percent of students proficient in mathematics. We interviewed district assessment personnel and used their expertise and knowledge and had them recommend schools where they knew formative assessment was practiced. Our efforts led to interviewing 108 of our top 250 districts, and we amassed a list of 504 potential participating schools. From this school list, we used our sorting methodology again to arrive at a preference list of schools and we had 122 schools sign up to participate in our project that fit our needs.
The methodology and procedures behind sample selection are incredibly important to projects, and contribute immensely to the end results of research questions. Our project shifted its focus early on and was able to successfully capture the schools and characteristics necessary for our results to be meaningful at the end of the day.

Author

• Kimberly Mundhenk, University of California - Davis