Co-Investigators: Kenneth A. Dodge, Amy B. Schulting

Purpose: Across the United States, high school dropout has emerged as a critical education policy problem. On average, one in four students who begin public high school in the United States will not graduate. Educators and policymakers have struggled to find the appropriate context and methods to intervene with students before they decide to drop out of high school. Most dropout prevention programs currently in place in American schools begin in adolescence, either toward the end of middle school or in the high school years. However, research suggests that the process of "becoming" a high school dropout can begin as early as elementary school. Researchers at Duke University argue that reducing elementary school truancy is a critical step in reducing middle and high school truancy, and ultimately high school dropout. In this study, researchers will develop, implement, and test the feasibility of a teacher-based truancy prevention program in elementary schools. Specifically, the truancy prevention program aims to:

1. support a collaborative relationship between parents and teachers and promote a positive teacher-child relationship;
2. provide teachers with timely data about a student's attendance patterns; and
3. train and supervise teachers in individualized interventions that are based on a particular family's reasons for truancy.

Project Activities: The core activity of this project is the iterative development of a truancy intervention program for the elementary grades that integrates universal teacher home visiting, timely monitoring of attendance patterns through an attendance information system, and identifying and addressing specific causes of early truancy. During the first year, the research team will develop and implement the intervention in two schools. During the second year, researchers will conduct a small, randomized pilot of the intervention in six North Carolina elementary schools to provide preliminary evidence of the effectiveness of the intervention in reducing elementary school truancy. Researchers will analyze pilot data and make final revisions to the intervention during the third year of the project.

Products: The main product of this project is a fully developed truancy intervention program for the second and third grades, including a teacher home visiting model, an attendance information system with daily attendance reports and user manual, and targeted protocols for identifying and addressing specific causes of truancy. Descriptions of the intervention, the development process, implementation, and preliminary outcomes will be shared in peer-reviewed publications.

Structured Abstract

Setting: Participating schools are located in a metropolitan school district in North Carolina.

Sample: The study population includes second- and third-grade teachers, their students enrolled during the 2012–13 and 2013–14 school years, and the students' parents.

The North Carolina metropolitan school district has approximately 32,500 students and is about 52 percent African American, 21 percent Hispanic, 21 percent white and 6 percent other. Approximately 63 percent of district students are from low-income families. This district had an average daily absenteeism rate of 6 percent over the past 3 years and about 10 percent of students have more than 10 absences per year, a common criterion for chronic absenteeism.

Intervention: The Early Truancy Prevention Project takes a comprehensive, multi-component approach to improving student attendance in the elementary grades. The first component is home-school relationship development. Second- and third-grade teachers will be trained and compensated to conduct a 30-minute home visit for each of their students at the beginning of the school year. The goal of this home visit is to establish effective home-school relationships. The second component is timely monitoring of attendance patterns. Researchers will develop and implement an electronic information system that provides teachers with immediate reports on each child's pattern of attendance over time so that teachers (as well as school administrators and the evaluation team) can identify a truancy or tardiness problem as soon as it develops. The final component involves identifying and targeting specific causes of absenteeism. Researchers will pilot an assessment to identify possible causes of absenteeism and a set of protocols designed to target specific causes. Based on attendance triggers, teachers will conduct attendance assessments and implement protocols targeted at a student's specific reasons for being absent. For example, teachers might work with school administrators to help solve specific transportation problems or work with the school nurse to address absences due to chronic or frequent health problems such as asthma, lice, or colds.

Research Design and Methods: During the initial year, researchers will develop the three intervention components and implement them in two North Carolina elementary schools. First, the research team will revise an existing kindergarten home visit protocol for use with families of second- and third-grade students. Researchers will also develop a web-based attendance system to document and track daily attendance, tardiness and early departures, truancy prevention efforts, and student and family responses. In addition, the team will refine and implement an Attendance Assessment Questionnaire that identifies causes of truancy among students and a set of targeted responses to specific causes of truancy. Researchers will also hold group discussions with teachers and collect written feedback in order to refine the intervention.

During the second year, researchers will conduct a small randomized pilot of the intervention in six schools. Second-and third-grade teachers will participate at each school; with one grade level participating in the intervention and the other grade serving as the control. Specifically, at three schools the second-grade teachers will implement the truancy prevention intervention and the third-grade teachers will conduct business-as-usual activities. At the other three schools, the third-grade teachers will implement the treatment and the second grade teachers will conduct business-as-usual activities. Across the six schools, approximately 24 classrooms will receive the intervention and about 24 classrooms will serve as controls.

During the third year, researchers will estimate the effect of the intervention on truancy by comparing the attendance record of students in the intervention group with those in the control group. This analysis, while lacking the power necessary for a definitive conclusion, should provide preliminary evidence of efficacy that could be used to justify conducting a full-scale randomized control trial. Researchers will also estimate the effect of the intervention on other student outcomes such as grades, disciplinary actions, and end-of-grade standardized test scores.

Control Condition: During the pilot test, control classrooms will conduct business-as-usual activities.

Key Measures: Student outcome measures include attendance (excused, unexcused, tardies, and early departures), grade retention/promotion, disciplinary infractions, grades (math, reading, science, etc.), special education referrals and end-of-grade standardized test scores. Researchers will also collect data on mediating mechanisms such as parent and teacher involvement, students' adjustment to school, student-teacher relationships, and teacher attitudes. During the development process and the pilot, researchers will also collect data on the fidelity and feasibility of implementation using teacher training evaluations, home visiting completion forms, home visit reflection forms, home visit observations, attendance assessment questionnaires, targeted intervention implementation forms, spot checks of attendance data accuracy, teacher interviews, teacher feedback surveys, and parent surveys.

Data Analytic Strategy: During the iterative development process, researchers will analyze data on the fidelity and feasibility of implementation. For the pilot, students are randomized at the classroom level, but observed at the student level. Researchers will use regression analyses (with errors clustered at the classroom level) to estimate the effect of the intervention on student-level outcomes. The cluster correction produces a credible estimate of the precision of the estimated treatment effect on student-level outcomes (such as total absences). Researchers will cross-check their findings by also analyzing the data using multi-level modeling. This will allow the team to estimate the proportion of variation in the outcome that is attributable to each level (e.g., individual student vs. teacher) of the design.

Products and Publications

Journal article, monograph, or newsletter

Cook, P.J., Dodge, K.A., Gifford, E.J., and Schulting, A. B. (2017). A New Program to Prevent Primary School Absenteeism: Results of a Pilot Study in Five Schools. Children and Youth Services Review, 82 , 262–270.