Improving Head Start attendance through behavioral science approaches

How to cite

Cojocaru, Laura. 2026. "Improving Head Start attendance through behavioral science approaches." AEA RCT Registry. AEARCTR-0018761. https://doi.org/10.1257/rct.18761-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Laura Cojocaru*

* Primary investigator

Registry ID
AEARCTR-0018761
DOI
https://doi.org/10.1257/rct.18761-1.0
First registered
Last updated
Status
On going
Trial dates
2025-10-06 – 2027-09-30
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/18761

Abstract

The Behavioral Interventions to Advance Self-Sufficiency-Next Generation (BIAS-NG) project is supported by the Office of Planning, Research, and Evaluation in the Administration for Children and Families, U.S. Department of Health and Human Services. BIAS-NG aims to make human services programs work better for the people receiving services by reshaping program processes using lessons from behavioral science, an interdisciplinary field that incorporates psychology, economics, and other social sciences to provide insight into how people process information, make decisions, and take action. BIAS-NG partners with state and local agencies to identify a challenge to address, investigate its possible causes, design an intervention informed by behavioral science to address the causes, and test the efficacy and cost efficiency of the intervention relative to status-quo service delivery.

The current study applies this approach to the challenge of chronic absenteeism in Head Start and Early Head Start (HS/EHS) programs in Wayne County, Michigan, and Franklin County, Ohio. Head Start is a U.S. federal program that provides early childhood education, health, and family support services to families with low incomes to help prepare children for school. While consistent attendance in early childhood education such as Head Start is linked to improved school readiness and long-term outcomes (McCoy et al., 2017; Yoshikawa et al., 2013), attendance is a challenge confronting many early childhood education programs (Connolly & Olson, 2012; Dubay & Holla, 2016; Rhoad-Drogalis & Justice, 2018.).

In partnership with the local programs, BIAS-NG developed a multi-component, behaviorally informed intervention designed to increase the salience of attendance, clarify the benefits of attendance and expectations for it, and support caregiver planning and self-monitoring. The intervention combines weekly attendance-focused electronic messages, monthly personalized attendance reports, and a home-based calendar tool.

The study uses a randomized controlled trial design, with families assigned to either the intervention or a business-as-usual control group. Impacts on attendance rate and chronic absenteeism are estimated using program administrative data on attendance. By rigorously testing this scalable strategy in real-world settings, the study assesses whether behaviorally informed supports can meaningfully improve attendance in HS/EHS programs.

Intervention

BIAS-NG’s attendance support intervention is designed to promote more consistent daily attendance in two HS/EHS programs by increasing the salience of attendance, strengthening caregivers’ understanding of how regular attendance supports children’s learning and development, and helping families plan for and track attendance. The participating programs serve preschool-aged children from families with low incomes and provide early center-based learning focused on school readiness, social-emotional development, and family engagement.

The intervention includes three integrated components delivered over approximately five months. First, caregivers are sent weekly messages through existing program communication apps. The messages highlight the developmental benefits of consistent attendance, reinforce positive norms related to attendance, and include occasional prompts for reflection or planning for regular attendance. Second, caregivers are sent monthly personalized attendance reports through postal mail. The reports show their child’s attendance to date. The reports are designed to increase awareness of attendance patterns and expectations. Third, families are given a take-home calendar that supports goal-setting, self-monitoring, and advance planning by encouraging caregivers and children to track attendance, anticipate potential challenges, and celebrate success. Together, these components draw on behavioral strategies such as personalization, feedback, salience, norm-setting, gamification, goal-setting, and planning prompts, and they build on prior evidence that behaviorally informed communication can improve early education attendance (for instance, Kalil, Mayer, & Gallegos, 2021.)

Experimental design

The study uses a randomized controlled trial (RCT) design to estimate the impact of a multi-component attendance support intervention on children’s attendance in two HS/EHS programs.

All eligible families with a child enrolled in the program during the study period will be included in the study sample. Families will be randomly assigned in a 1:1 ratio to either a treatment group or a control group. Randomization occurs at the family level so that all children within the same family receive the same assignment, reducing the risk of confusion or spillovers within households. Randomization will be conducted separately within each program site. Randomization is stratified by center, ensuring balance in the number of treatment and control families within each center.

Random assignment is conducted on a monthly basis through February. At the beginning of each month, programs generate an updated file of newly enrolled eligible families based on administrative data covering enrollments through the end of the previous month. These newly eligible families are then randomized using the study’s random assignment system and added to their assigned study group. Families already randomized remain in their original group.

Families assigned to the treatment group are given the behavioral intervention that includes: (1) weekly attendance-focused messages delivered through the program’s existing communication platform, (2) monthly personalized attendance reports summarizing the child’s attendance delivered by postal mail, and (3) a take-home calendar tool to support attendance goal-setting and tracking.

Children enrolled in partnership centers (HS/EHS slots located within partner child care settings) are not given the calendar component because their operating schedules do not align with the standard program calendar. These families are still given the messaging and attendance report components.

Families assigned to the control group continue to receive business-as-usual services and communications from the program (such as routine announcements and standard attendance-related outreach) but are not given the study’s attendance-focused messages, personalized attendance reports, or calendar tool.

Outcomes will be measured using administrative attendance records collected by the programs for the duration of the intervention.

One or both partner sites may continue to implement the intervention for the treatment group after the conclusion of the project.

Design details

Randomization method
MDRC random assignment web tool
Randomization unit
Families were the unit of randomization. The randomization was also stratified by Head Start center.
Clusters
1,500 Families
Observations
1,650 children
Arms
Intervention group: 825 children, Control group: 825 children
Primary outcomes
The primary outcomes focus on children’s attendance during the intervention period, measured using administrative attendance records. Attendance rate: The proportion of scheduled program days that a child is marked present during the observation period. Chronic absenteeism: An indicator of whether a child is chronically absent, defined as missing more than 10 percent of scheduled program days during the observation period. We will consider the study period to be the four months during which the intervention was fully implemented – from November 1, 2025, through February 2026. These outcomes will be examined over that study period and, in additional analyses, by month to assess how impacts evolve over time. We will also analyze whether impacts persist three months after the study period ends and either the intervention is completely ended or the site(s) take over its implementation.
Secondary outcomes
Average number of days missed per month: The average number of days a child has missed per month, calculated for months when they were fully enrolled in the program. Attendance rate excluding illness-related absences (if possible to calculate): Proportion of scheduled program days the child was marked present, calculated after excluding days for which caregiver-reported illness was recorded as the reason for absence. This outcome is exploratory and will be interpreted cautiously, as self-reported absence reasons may be measured with error and could be influenced by the intervention. Expulsion rate (if possible to calculate): Proportion of children who were formally expelled during the observation period. Expulsion is a binary outcome defined as one for children who were expelled from the program – usually due to poor attendance – and zero otherwise. Drop-out rate: Proportion of children who drop out of the program during the observation period, where drop out is a binary outcome coded as one for children who voluntarily left the program, and zero otherwise. Caregiver knowledge and attitudes related to attendance (survey-based, if fielded): Perceived acceptable number of absences in a typical month, calculated as the average number of absences indicated in response to the question “In a typical month, about how many days do you think it is OK for a child to miss Head Start?”  Knowledge of whether the program has minimum attendance expectations, calculated as the percentage of caregivers that respond “Yes” to the question “To your knowledge, is there a minimum number of days per month that Head Start expects children to attend?”  Awareness of possible program actions related to low attendance, calculated as the percentage of caregivers who respond “Yes” or “Maybe” to the question “Are you aware of any actions that Head Start might take if a child consistently misses school?”  Beliefs about whether consistent attendance supports child learning and development, measured using the question: “Do you believe that missing two days or more per month at Head Start negatively affects your child’s learning or development?”  The primary indicator will be the percentage of caregivers who answer “Yes, a lot.” We will also provide a descriptive comparison across all response options: Yes, a lot Yes, a little Maybe, but it depends on the reason for the absence Only if it happens frequently No, not really I’m not sure Beliefs about whether children benefit more from daily attendance versus staying home, calculated as the percentage of caregivers who select “A child benefits more from attending Head Start every day.” In response to the question “Which of the following statements best matches your views?”  Satisfaction with the child’s current attendance level, measured as the percentage of caregivers who answered "No" to the question “Did your child attend Head Start as often as you wanted this school year (starting in September)?” Perceived norms regarding the child’s attendance relative to classmates, described by comparing the percentage of caregivers who believe their child attends a lot more, more, about the same, fewer, or a lot fewer days than other children, or that do not know. We will not be verifying how this corresponds to their child’s attendance. However, higher percentages of caregivers reporting “a lot more” or “more” would indicate a tendency to perceive their child as attending more than peers, highlighting potential overestimation of attendance relative to classmates. Caregiver-reported barriers to attendance (survey-based, if fielded): The calculation will be based on caregivers’ report of how often specific factors contributed to their child’s absences from HS/EHS from September through March, using a frequency scale ranging from Never (0 days) to Very often (>15 days). Factors include illness, transportation problems, caregiver availability, child reluctance, family preferences, weather, scheduling confusion, and caregiving arrangements. Outcomes will be summarized as the percentage of caregivers selecting each frequency category for each barrier. We will also calculate the proportion of children affected by each barrier at least once, noting that responses reflect caregiver perceptions rather than verified reasons for absences.

Cited by

No citation data available for this registration yet.

Keywords