OVERVIEW

Even Start is a family-focused intervention program designed to break the intergenerational cycle of poverty and low literacy. The program has three goals: (a) to "help parents improve [family] literacy or basic education skills," (b) to "help parents become full-time partners in educating their children," and (c) to "assist children in reaching their full potential as learners" (U.S. Department of Education, 1998). Even Start components include adult education and literacy, parenting education, early childhood education, and various support services (e.g., transportation, child care, counseling). Experimental evaluations show that participation in Even Start has produced certain modest positive impacts for parents and children. Parents in Even Start were more likely to attain their GRE than control parents; however, program parents did not improve literacy, parenting or personal skills. Compared to their control group peers, Even Start children performed significantly higher on one measure of school readiness; however, other measures of school readiness found no difference between the groups. Finally, older Even Start participants (age five at program debut) had significantly fewer behavior problems than the control group participants; this decrease was not found, however, for younger participants (0-4 at program debut).

DESCRIPTION OF PROGRAM

Target population: Parents and children up to age 8 from low-income families. Parents must be eligible to participate in an adult education program under the Adult Education Act

Authorized in 1989, Even Start is a family-focused intervention and has three core components that address its goals. The first component is adult education and literacy. Included in this are Adult Basic Education (ABE), English as a Second Language (ESL) and training for obtaining the General Education Development (GED) Certificate. The second component is parenting education, which includes instruction for parents on supporting their child's education. The third component is early childhood education, which includes direct intervention to prepare children for school. All Even Start sites are required to address each of these three core services, however the specific activities and services in each component is left up to the discretion of the individual sites. This allows a site to tailor services to the needs of a particular population and focus on specific outcomes within that population. Even Start sites are also required to provide some core services through joint sessions with parents and children.

In addition to the three core components, Even Start sites may provide support services. These services may include transportation, child care, nutrition services, screening and referral for chemical dependency, referrals for mental health and counseling, and job referrals. Even Start also includes an evaluation component.

Even Start components are coordinated with existing service providers in communities, when possible, to avoid duplication of services.

· Number of Children/Teens in Program: There are approximately 800 sites nationwide serving approximately one million parents and children. The average site serves approximately 54 families.
· Length: Most sites serve families with children from birth through age 8, however sites may choose to narrow their focus to a smaller age range.
· Intensity: Families participating in Even Start are required to participate in activities in all three components of the program. Each Even Start program is expected to provide at least 60 hours of adult education, 20 hours of parenting education and 65 hours of early childhood education per month. On average, parents are involved in 95 hours of adult education and 35 hours of parenting education services per year. Children spend an average of 15.5 hours per month in early childhood education provided by Even Start programs. A majority of families only participate for one year; less than a quarter of families participate for more than two years.
· Service Delivery Mode: Varies from site to site.

EVALUATION(S) OF PROGRAM

Three four-year national evaluations of Even Start have been conducted by the U.S. Department of Education. The first evaluation was conducted from 1989 through 1993, the second from 1993 through 1997, and the third from 1997 through 2001. We will focus on the first and third evaluations because they utilized random (experimental) assignment (U.S. Department of Education, 1995, 2003).
 

U.S. Department of Education. (1995). National evaluation of the Even Start Family Literacy Program: Final report. Washington, DC: U.S. Department of Education, Planning and Evaluation Service.
 

Evaluated population: 200 program families were from all ten sites while the 100 control families were from five of the ten sites. Children in the in-depth study ranged in age from birth through 5 years of age with most children being between ages 3 and 5. The sample was ethnically diverse; some sites served only black families, some only white, and others mixed. Hispanics, Native Americans and Afghanis were also represented in the sample.

Approach: The first evaluation used the National Evaluation Information System (NEIS), which is used to collect common data from all Even Start sites (U.S. Department of Education, 1995). These data were used to examine implementation during Even Start's first four years of operation, as well as effects on the first two cohorts of participants. The evaluation also included an in-depth study of ten Even Start sites. This in-depth study included random assignment of families to treatment and control groups in order to examine the effects of Even Start on participants.

The NEIS component of the Even Start evaluation included data from 340 sites. The program implementation evaluation found that the program was serving its target population of low income, needy families with children less than 8 years of age. Further, the number of families participating in Even Start increased over time from 2,500 families served in 76 sites in the 1989-90 school year to 20,000 families served in 340 sites in the 1992-93 school year.

To examine the effect of Even Start on families, the evaluation drew on NEIS data and in-depth study data. In this summary, we focus on the experimental findings from the in-depth study, which consisted of ten Even Start sites. In five of the ten sites, participants were randomly assigned to a control group or an experimental group; in the other five sites, all of the families participated in Even Start (so they were all in the experimental group). The Families in the in-depth study were recruited in the summer and fall of 1991, the program occurred in 1991 and 1992, and post-test data were collected in the summer of 1992 and spring of 1993.

Results: Some significant effects were found for children. Children in the Even Start group exhibited higher gains than control participants on the Preschool Inventory (PSI; a measure of school readiness), however these differences faded after 18 months. Even Start children were also more likely to participate in early childhood education programs (including those offered through Even Start) than control children. The researchers found no significant differences between children in the Even Start and comparison groups on the Peabody Picture Vocabulary Test (PPVT) or the Child's Emergent Literacy Test. In the area of parent literacy, a greater percentage of Even Start parents received their GED when compared to control parents (22.4 percent vs. 5.7 percent). However, on the reading survey component of the Comprehensive Adult Student Assessment System (CASAS; a measure of adult functional literacy in reading), program parents did not show significant differences from control parents. Likewise, there were no significant difference between program and control participants on measures of parental reading and writing behaviors in the home. On measures of parenting skills, Even Start participants did not experience differences when compared to control participants. The evaluation found no significant differences on measures of parents' personal skills (e.g., sense of well-being, sense of control over life), parent-child reading behavior, or parental educational expectations of children. In regard to the home learning environment, only one measure indicated a significant difference: Even Start parents reported significantly higher numbers of reading materials in the home than control participants.

Overall, the results were mixed. It should be noted that there was substantial variation in level of program intensity and in the degree of family involvement.
 

U.S. Department of Education. (2003). Third national Even Start evaluation: Program impacts and implications for improvement. Washington, DC: U.S. Department of Education, Planning and Evaluation Service.

Evaluated population: The Experimental Design Study (EDS) sample consisted of 463 families; about two-thirds of the sample (309 families) was randomly assigned to an experimental group and one-third (154 families) to a control group. Children in the program were between birth and 5 years of age with a majority of children between 3 and 5 years of age.

Approach: Similar to the first national evaluation of Even Start, the third national evaluation had two components (U.S. Department of Education, 2003). The first was data collected through the Even Start Performance Information Reporting System (ESPIRS), which collects common data from all Even Start sites. The second component was the Experimental Design Study (EDS). In this summary, we focus on findings from the EDS, which included 18 sites. These sites were not representative of the Even Start population because Hispanic families in urban areas were over-represented. Participants in the experimental group began the Even Start program immediately while participants in the control group were not allowed to begin the Even Start program for one year. Data were collected at pre-intervention, post-intervention one year later, and in a one-year follow-up. Results of the one-year follow-up have not been released.

Results: Results of the study showed very few significant differences between the experimental and control groups. On measures of literacy, experimental group parents and children did not show significantly better performance than control group parents and children. Further, teacher ratings of children's social skills did not differ significantly between the control and experimental groups. There were also no differences in school records in terms of tardiness, absences or special education placement. The only significant difference between the control and experimental groups were teacher ratings of behavior problems using the Social Skills Rating System (SSRS). Teachers rated elementary school age experimental group children as having fewer behavior problems than control group children. However, this difference was not significant for preschool children, who were rated using the SSRS by preschool teachers.

SOURCES FOR MORE INFORMATION
 

Link to program curriculum: http://www.evenstart.org/index.shtml 

References:

U.S. Department of Education. (1993). National evaluation of the Even Start Family Literacy Program: Report on effectiveness. Washington, DC: U.S. Department of Education, Office of Policy and Planning.

U.S. Department of Education. (1995). National evaluation of the Even Start Family Literacy Program: Final report. Washington, DC: U.S. Department of Education, Planning and Evaluation Service.

U.S. Department of Education. (1998). Even Start: Evidence from the past and a look to the future. Washington, DC: U.S. Department of Education, Planning and Evaluation Service.

U.S. Department of Education. (2003). Third national Even Start evaluation: Program impacts and implications for improvement. Washington, DC: U.S. Department of Education, Planning and Evaluation Service.

Wasik, B.H. (2003, April). Family literacy interventions. Paper presented at the Biennial Meeting of the Society for Research in Child Development, Tampa, FL.

Program also discussed in the following Child Trends publication(s):

Child Trends. (2001). School readiness: Helping communities get children ready for school and schools ready for children (Research brief). Washington, DC: Child Trends.

Halle, T., Zaff, J., Calkins, J., & Margie, N. G. (2000). Background for community-level work on school readiness: A review of definitions, assessments, and investment strategies. Part II: Reviewing the literature on contributing factors to school readiness. Washington, DC: Child Trends, Inc.

SUMMARY & CATEGORIZATION

Program categorized in this guide according to the following:

Evaluated participant ages: 0-5 / Program age ranges in the Guide: 0-5

Program components: child care/early childhood education, clinic/provider-based, parent or family component

Program information last updated 3/14/07

© Child Trends 2003