Design, Sites, and Data Sources
The Jumpstart Foster Grandparent Program evaluation will conduct a rigorous study in 11 Los Angeles early childhood education centers spread across 6 agencies during the 2017-18 program year. Random assignment will be used to assign children at participating centers to one of three types of classrooms: (1) a classroom where all children receive Jumpstart sessions and CCT during nonsession time, (2) a classroom that will receive only CCT and no Jumpstart sessions, and (3) a classroom that will not receive volunteer services. Through this design, it will be possible to evaluate the effect of the full Jumpstart model as well as the isolated effect of CCT only. To meet the sample size requirements, the study will also include classrooms where, because random assignment is not possible, matching methods will be used to choose a comparison group of children. Through this process, there will be minimal changes to the current program’s implementation, meaning that the study will be able to evaluate the effects of the Jumpstart program as it is typically implemented in authentic classroom settings.
Furthermore, the evaluation aims to minimize the data collection burden on participating centers by relying on data that is already collected. Specifically, children’s language/literacy and social-emotional development will be measured using the Desired Results Developmental Profile Preschool (DRDP-PS). The children’s receipt of session and CCT hours will be measured using attendance records and time sheets. Lastly, program implementation fidelity and quality will be assessed using Jumpstart’s monitoring tools, such as volunteer training trackers, volunteer surveys, and volunteer timesheets, supplemented with qualitative interviews with volunteers conducted by MDRC.