MDRC Evaluates Jumpstart’s Foster Grandparent Program in California


The hallmark of Jumpstart’s direct service model is a curriculum that volunteers deliver during twice-weekly sessions that focus on preschool children’s language, literacy, and social-emotional development. All volunteers also spend time outside of those sessions in order to provide additional support to preschool children in their classrooms.

While 96% of Jumpstart’s volunteers are college-age students, from 2005-2019, Jumpstart also engaged older adult volunteers from the community (Community Corps members or Foster Grandparents). From 2011 until 2019, Jumpstart offered a Foster Grandparent Program (FGP) in Compton and Los Angeles, California. Across two of the operating years of the FGP model (2016-2017 and 2017-2018), Jumpstart received resources from the Corporation for National and Community Service (CNCS) to work with the research firm MDRC to plan for and then execute a rigorous study to examine the impact of this program model.

While, like the college model, the FGP model offered Jumpstart’s core programming and curriculum, there are notable differences that limit the applicability of the findings from this study. Some examples follow.

  • Perhaps most significantly, Jumpstart is now implementing a revised curriculum (beginning in 2018-19 throughout our network of 700+ classrooms) with a narrowed focus on oral language and emotion understanding. The prior curriculum was in use by the FGP teams during the study.
  • There are administrative and logistical differences between the college and FGP models that affected the program’s implementation. For example, in the FGP teams, there was a more fluid distribution of roles and rotating leadership among Foster Grandparent team members versus a dedicated team leader for the college model. A study of our college model (Dwyer, 2012) found that of all adults in a classroom, team leaders offer children the most oral language support. A lack of a team leader in the FGP may have implications for children’s language and literacy gains.
  • For Foster Grandparents, all interactions occurred in-person or via phone rather than virtually. This meant that member applications, timesheets, surveys, feedback, and other data were collected in paper form and required entry into systems. This required staff to dedicate more time to administrative tasks and less time supporting levers of quality (e.g., offering additional coaching visits) than in our college model.

Although the results of this study were inconclusive and the FGP program differed in significant ways from our foundational college program, Jumpstart is taking seriously the findings that do apply because of our deep commitment to continuous improvement and to providing the highest quality of programming for the children we served. We share some of the learnings from the study here and our actions in response.

The Results

The findings from this study – which can be found in this report released by MDRC – show the following:

Jumpstart Foster Grandparents:

  • had consistently high attendance at training sessions and reported feeling prepared for providing services to children.
  • demonstrated a strong commitment to the program, with some volunteers exceeding their service hour requirement of 15 hours a week.
  • implemented Jumpstart sessions regularly as intended and used a variety of intentional instructional practices.
  • devoted a high number of hours to Child Centered Time (CCT; time outside of the Jumpstart sessions where volunteers interact with children one-on-one and contribute to their language and social-emotional development within the setting of teacher-led instructional time).


  • were supported by an increased number of adults in Jumpstart classrooms.
  • who received the full Jumpstart model (Jumpstart sessions plus CCT) made pre- to post-intervention gains in the Language and Literacy Development and Social-Emotional Development domains of the Desired Results Development Profile (DRDP) assessment. Ultimately, children receiving the full Jumpstart model performed the same as comparison children on both domains.
  • who received CCT-only made numerically larger gains than comparison children. The difference between CCT-only and comparison children’s gains was moderate (in terms of the effect size), but not statistically significant.

Data Collection Challenges and Limitations

The MDRC study used a rigorous random assignment design* to evaluate the impact of the FGP program on children. However, as the report from MDRC highlights, data collection was challenging. Sample sizes were small (primarily driven by the mobility of the population of children being studied), the amount of time that volunteers spent one-on-one with children was difficult to measure, the outcome assessment used for children may not accurately measure children’s true level of development, and the biases inherent in self-report (which was used to measure the frequency of teachers’ language and literacy practices) were hard to parse. Additionally, in terms of the design, children were randomly assigned to classrooms, but teachers were not. Therefore, there could be systematic differences in teacher characteristics or quality across the classrooms that were not accounted for in the study design, analyses, or results.

Application of Learnings

While the Community Corps program differed in meaningful ways from Jumpstart’s foundational college program, Jumpstart is applying the findings from this study more broadly.

In response to data collection challenges:

  • Per MDRC’s suggestion, Jumpstart will work to include direct observations of volunteers instead of relying exclusively on self-report in future rigorous studies.
  • Jumpstart is working to refine its quality assurance processes and to intentionally link observation data with child outcomes.

In terms of program elements:

  • Jumpstart is already responding to the CCT-only results by offering a more intentional version of CCT service across all sites this year (2019-2020) called Individual Classroom Service (ICS). Newly developed ICS tools provide Jumpstart volunteers with specific strategies for engaging children in conversation and activity ideas for building vocabulary and comprehension skills for individual children related to session storybooks and unit themes (e.g., family, friends, and wind and water).

In summer 2019, Jumpstart discontinued its Foster Grandparent program to focus its resources on its College Corps program, which strategically aligns with Jumpstart’s goal of training volunteers who may later choose to become early childhood educators, and to build on our long-standing partnerships with institutions of higher education.


*Random assignment is widely accepted as the “gold standard” of research designs intended to determine what effect a program has on the outcomes of its participants. 

No comments yet.

Leave a comment

Your email address will not be published.

Together, we can help all children build the key language and literacy skills they need to take on the world.

Donate Now