Methodology

Great Start Readiness Program

Methodology for Academic Impacts

We proposed a quasi-experimental design study where the future academic performance of the GSRP participants will be compared to the performance of the waitlist children. The waitlist children are defined as GSRP applicants who complete the application process but never attend a publicly funded preschool program.

Given the limited study period, the academic outcomes to be focused include:

The rationale is that if GSRP is effective, GSRP children should achieve positive learning outcomes and sustain those positive outcomes as they get older compared to the waitlist children. The waitlist children form a good comparison group based on quasi-experimental design standards because they are more like the GSRP families (in terms of being aware of and motivated to enroll) than a matched control group exclusively based on income and demographic characteristics. Access to such data is currently pending for approval by the State.

Methodology for Economic Impacts

We proposed to translate the academic impacts into dollar amounts in terms of ISD and the state saving on special education, grade retention, and absenteeism expenditures. We are currently exploring options with Michigan Department of Education (MDE) to collect additional information on local grantee’s in-kind contributions (i.e., busing) and other funding sources (i.e., Title I, food programs, school improvement grants) that support GSRP classrooms to get a more accurate picture of true cost.

Overview of the Available Data

Mainly utilizing the secondary data for analysis, our goal is to identify a way to retrieve data from different sources and connect them in an efficient way to maximize data usage. Given the natural layers of data structure, our design of the analysis will account for the nested effects.

The highest level for data collection is at the ISDs, focusing on the funding amount and utilization. Because ISDs select local subrecipients to manage regional sites, the second level for analysis will be how different types of the subrecipients, the expenditures and the serving capacities between Head Start and GSRP classrooms, might relate to the outcomes.

Subrecipients usually manage multiple sites, with each site having a valid childcare licensing number. At the site level, we will examine how varying site types, geographic locations, and curriculum and assessment tools used for the sites might show different results.

At the classroom level, data around program quality assessments (i.e., PQA-R, CLASS), teacher credentials and salaries, and classroom options (school-day, half-day and Head Start blend) will be used for analysis.

For data related to students’ characters and academic performance, a data sharing agreement between MSU’s Community Evaluation and Research Collaborative (CERC) at University Outreach and Engagement and Michigan’s Center for Educational Performance and Information (CEPI) has been executed in October 2018.

Notably, because the State collects child data from the whole program site, not individualized by classrooms, all the classroom data will be compiled at the program site for analysis.

An overview of the data availability at different levels can be found in this illustration.