School Readiness and Summer Slide: Supporting Parents in the Summer Before Kindergarten

Ready4K is a text-messaging program for parents of newborns to third graders designed to help them support their children’s development. The program disseminates research-based text messages that show parents how to create everyday learning opportunities in fun and easy ways that enhance their typical home behaviors. The text messages integrate best-practice instruction with existing family routines and align with state educational standards. The program is available in multiple languages and was developed with particular attention to cultural appropriateness in diverse settings. Rigorous evidence, from an RCT study, demonstrates that Ready4K programming leads to student learning gains as well as changes in parent practices (York & Loeb 2014). A follow-up RCT study with kindergarten students documented larger learning gains when messages were tailored to students’ instructional levels (Doss, Fable, Loeb & York 2017).

The current study builds on the existing evidence to explore the effectiveness of a text message-based developmental curriculum in the summer months before kindergarten entry. This study will augment previous research to explore the efficacy of a summer supplement to the Ready4K curriculum, particularly targeted at addressing the issue of summer slide among disadvantaged children in their out-of-school time with the aim of boosting children’s school readiness.

Working with the program developer and state partners in Wisconsin, this study will roll out in the field in the 2018-2019 academic year. The Read to Lead Council and others in Wisconsin are particularly interested in the summer months preceding kindergarten entry as a critical period for school readiness. Accordingly, the intervention will include supplemental summer content to test the effectiveness of programming—intended to mitigate summer slide and remediate school-entry skill gaps—provided to parents of rising kindergarten students in the months before school starts. There are three treatment conditions to which parents will be randomly assigned:
1. Ready4K over the course of their child’s 9-month preschool academic year, September 2018 to May 2019
2. Ready4K plus supplemental summer content in the three months before their child’s kindergarten entry, September 2018 to August 2019
3. Ready4K plus supplemental summer content that is tailored to the family’s specific geographic area, September 2018 to August 2019

2018-08-16

2019-09-15

The primary outcome measure is performance on kindergarten-entry assessments–the Phonological Awareness Literacy Screening, Kindergarten (PALS-K) and the Measures of Academic Progress (MAP)–administered in the fall of the kindergarten year, fall of 2019 for the relevant study cohort. PALS-K is administered at kindergarten entry as an early literacy skills screening tool, and was the required literacy screener in Wisconsin from 2012–13 until 2015–16. While school districts are now allowed to select their kindergarten readiness screener, PALS-K is still the most widely used in the state. One participating district uses the NWEA MAP assessments as their kindergarten-entry skills assessment.
The PALS-K screener focuses on fundamental early literacy skills, including phonological awareness, alphabet knowledge, knowledge of letter sounds, spelling, concept of word, and word recognition in isolation. Intended to gauge children’s progress in meeting developmental milestones in literacy acquisition, the research team that developed PALS-K, at the University of Virginia, has provided evidence of the instrument’s reliability and construct, concurrent, and predictive validity. MAP assessments, developed by NWEA with a substantial body of evidence on reliability and validity, are also widely used across the United States. MAP includes a reading fluency screener as well as developmentally appropriate assessments in early literacy and numeracy.

For future analyses, the Wisconsin Forward Exam, initiated in the 2015–16 school year, is administered online in the spring of each school year and covers third through eighth grades in English language arts and mathematics, fourth and eighth grades in science, and grades four, eight, and ten in social studies. The four-year old cohort participating in the study will participate in the third grade ELA exam in the 2022–23 school year.
Another secondary outcome measure are third-grade test scores of children. This study will allow for longer-term follow-up of study participants as children progress into the later grade levels and are tested in mathematics and reading on standardized state assessments.

This study employs an RCT design to assess the causal impact of Ready4K summer supplements on children’s school readiness. The main research questions of interest are:
Does parent participation in a text-messaging program, designed to support children’s healthy development and early skill-building, in the summer months before kindergarten impact children’s kindergarten readiness?
Does a tailored summer program, with specific content for the geographic area where the family resides, affect kindergarten readiness differently than a general summer program?
Is there variation in program impact by characteristics of the preschool setting or characteristics of the child and family?
Parents of four-year old children in participating early childhood program sites, 4K and Head Start, will be randomly assigned to one of the three treatment conditions with equal probability within each site. In this block-randomized design, early childhood program sites are the blocks, and individual parents within sites will be assigned to each treatment condition (one-third of the within-site sample in each condition). For the purposes of answering the first research question, the two summer supplement treatment groups will be pooled.
This effort and accompanying study will be deployed in Wisconsin’s public pre-kindergarten program, 4K, and Head Start sites in three communities across the state. These are among the largest school districts in the state, facilitating large and diverse samples. Because the preschool programs in question are subsidized, children come predominantly from low-income families, particularly in Head Start programs for which eligibility is conditioned on family disadvantage. The largest school district in particular, located in an urban area, has a racially, ethnically, and linguistically diverse student population. Ready4K programming, including the summer supplement, is available in English and Spanish and has been developed with cultural appropriateness in diverse settings in mind. For the purposes of this project, the research team will also work with Ready4K to make the program accessible to families in the study districts.
With assumptions about the proportion of the four-year old cohort that will matriculate in kindergarten in the fall of 2019 (80 percent), and the parent opt-in rate (75 percent), we arrive at a sample size of approximately 1,000 families in the three communities, with approximately one-third of participating parents assigned to each treatment condition. An early childhood program site in this context includes schools, private childcare centers with state-funded 4K slots, community-based centers, and Head Start centers.

Randomization conducted by the research team using a computer program.

The unit of randomization is the family. Randomization will be stratified by classroom, so that in a particular classroom, approximately ⅓ of participating families are assigned to each treatment arm of the intervention. This design was chosen to mitigate an unequal distribution of students to a particular treatment arm with similar characteristics, such as home language or socioeconomic status.

No

1,100 students

1,100 students

366 students receiving Ready4K only throughout the academic year
366 students receiving Ready4K standard summer content
366 students receiving Ready4K geographically-tailored summer content

Varying the assumptions on power analysis calculations suggests that the approximate study sample size of 1,100, across 43 sites, would provide sufficient power to detect an effect size of 0.089–0.125 standard deviations. All power analysis calculations were conducted with the PowerUp! Program. These minimum detectable effects are consistent with effects of early childhood and parenting programs in the existing literature on similar outcomes of interest. The baseline assumptions for these calculations are power (1-β) of 0.80, an α, or significance level, of 0.05 (or the conventional 4-1 tradeoff in Type II and Type I errors), and a two-sided test of between-group differences in means. Power analysis calculations pertain primarily to the first research question, or the main treatment contrast of interest (supplemental summer content versus no supplemental summer content), so the proportion of Level-1 units randomized to treatment is two-thirds. One-third of participating parents are assigned to the control group and do not receive supplemental summer programming. The block-randomized design, in which parents are assigned to treatment conditions within sites was also taken into account, as was within-site clustering in the outcome measure. The power analysis is based on the approximately 43 sites with approximately 26 participating parents within each site. Finally, these power calculations are based on a range of assumptions about covariates with the inclusion of five to 10 baseline child and family covariates (R2 ranging from 0.5 to 0.75), including a developmental assessment administered in the spring of the pre-K year that likely explains a significant portion of the variance in the kindergarten readiness assessment. Moreover, inclusion of site fixed effects also contributes to the precision of the estimation.