A r c h i v e d  I n f o r m a t i o n

Background
Characteristics of Even Start Participants
Implementation of Even Start Projects
Participation in Even Start
Effects of Even Start (Short-Term)
Cost of Even Start

Background

The Even Start Family Literacy Program was authorized by the Elementary and Secondary Education Act of 1965, as amended by the Hawkins-Stafford Elementary and Secondary School Improvement Amendments of 1988, Part B of Chapter 1 of Title I (P.L. 100-297). The Even Start program is intended to:

...improve the educational opportunities of the Nation's children and adults by integrating early childhood education and adult education for parents into a unified program....The program shall be implemented through cooperative projects that build on existing community resources to create a new range of services. (P.L. 100-297, Sec. 1051).

To be eligible for Even Start, a family must have an adult who is eligible to participate in an adult education program under the Adult Education Act, and who is a parent of a child less than eight years of age who lives in a Chapter 1 elementary school attendance area. Even Start projects must provide participating families with an integrated program of early childhood education, adult basic education, and parenting education. The program's design is based on the notion that these components build on each other and that families need to receive all three services, not just one or two, in order to effect lasting change and improve children's school success.

Even Start began as a demonstration program administered by the U.S. Department of Education (ED) that provided school districts with discretionary grants for family literacy projects. In 1992, the program, while remaining a competitive discretionary grant program, became administered primarily by the States, with set-aside amounts for Federally-administered grants to Indian tribes and tribal organizations and for Migrant Even Start projects. Each State's share of Even Start funding is based on its proportion of Chapter 1 Basic Grants funding. States hold grant competitions and make subgrant awards to local school districts or nonprofit organizations. The statute specifies that each Even Start subgrantee must receive a minimum of $75,000 per year.

The following information, from the National Evaluation of the Even Start Family Literacy Program, is based on three years of data collected from the first 123 projects funded under Even Start--the two cohorts of projects that started in FY 1989 and FY 1990. These projects have participated in a detailed evaluation of their programs and services and have now submitted data on their projects for 1989-90, 1990-91, and 1991-92 program years.

Characteristics of Even Start Participants

Even Start projects are serving the intended population. All of the participating Even Start households had at least one child between birth and age eight, 77 percent of the adults who participated in Even Start core services did not complete high school, and 69 percent of Even Start families had total annual income under $10,000. The Even Start population can be further described as follows:

• 48 percent of Even Start families describe themselves as couples with children, 39 percent are single parent households, and 13 percent have extended families or other living arrangements.

• 47 percent of Even Start families report job wages as their primary source of financial support, while 52 percent report that government assistance is their primary source of support.

• Most adults in Even Start are between the ages of 22-29 (47 percent) or 30-39 (31 percent).

• 45 percent of Even Start adults are white, 27 percent are African-American, 18 percent are Hispanic, 6 percent are Native American, and 3 percent are Asian or Pacific Islander.

• English is the primary language for 72 percent of Even Start adults; Spanish is the primary language for 22 percent.

• Nine percent of the children served by Even Start were identified as having a disability.

• Even Start parents show high rates of depression--45 percent of parents in the In-Depth Study were rated as clinically depressed on the CES-D rating scale.

Implementation of Even Start Projects

Core Services Delivered Through Even Start

• Parenting Education Services. Even Start projects provided a wide range of services to help parents raise their children. For example, projects helped families make use of services provided by other social agencies, emphasized parents' role in the education of their children, oriented parents and children to school routines, furnished information about child development, trained parents in child behavior management, worked to build parental self- esteem, and instructed parents in life skills and in health and nutrition. Each of these different types of parenting education was provided by 90 percent or more of the Even Start projects. Depending on the specific service, 34 to 55 percent of the Even Start projects provided parenting education directly, 36 to 54 percent shared provision with a cooperating agency, and 5 to 10 percent delegated provision of parenting education to a cooperating agency.

• Adult Education Services. All projects provided adult education services, although not all arranged for education at the basic education level, secondary education level, and English as a second language education. Almost 100 percent of the projects reported that they provided services to prepare adults to attain a GED; 86 percent provided services in adult basic education. Instruction in English as a second language was provided by 59 percent of the projects. The locus of responsibility for providing adult education services differs from that of parenting education. About 30 percent of the projects provided adult education services directly by Even Start staff, another 30 percent of the projects provided services through cooperating agencies, and about 25 percent of the projects provided services jointly by Even Start and cooperating agency staff.

• Early Childhood Education Services. Children in Even Start projects were provided with a range of early childhood education services. Three different preschool options were used, with many projects using combinations: (1) 72 percent of the projects enrolled some of their children in Head Start, (2) 46 percent of the projects enrolled some of their children in a basic Chapter 1-funded pre-K program, and (3) 93 percent of the projects provided some other preschool option. For children old enough to enter the public schools, most Even Start projects participated in joint planning activities with the public schools, in addition to providing comprehensive family services such as afterschool programs or other education programs for the children as well as services for parents. Hence, 85 percent of the projects included kindergarten as an Even Start service, and 74 percent of the projects provided early childhood education services to children under eight years of age who were in primary grades, again through the vehicle of joint planning with the public schools and comprehensive family services.

  As would be expected, all Head Start and Chapter 1 pre-K services were provided by cooperating agencies. About 38 percent of the "other preschool" services were provided directly by Even Start staff. This distribution is not surprising given the high cost of early childhood education services and their general availability through cooperating agencies and the public schools.

Adult/Child Services Delivered Through Even Start

Almost all of the Even Start projects (more than 90 percent) reported that they delivered a wide range of core services to parents and children together. This is an important part of the Even Start model in that it impresses on parents that they are a key to their child's education and that parent/child interactions are critical for child development. Examples of services delivered to adults and children together include reading and story telling, developing readiness skills, social development and play, development of gross motor skills, working with numbers, arts/crafts, and health/nutrition.

Support Services Delivered Through Even Start

Even Start projects used funds to provide many different kinds of support services designed to enable families to participate in the core services. Comprehensive support services remove barriers that, if unattended, restrict a family's ability to receive instructional and educational services. Several types of support services were provided by 85 percent or more of the projects including transportation, referrals for employment services, mental health services, family advocacy assistance with government agencies, child care, meals, counseling, nutrition assistance, and health care assistance.

Across all types of support services, Even Start projects provided 22 percent directly, cooperating agencies provided 27 percent directly, and Even Start and cooperating agencies jointly provided another 28 percent. These findings suggest that Even Start projects did, as planned, obtain many support services from existing providers and stepped in to provide more immediate support services to enable families to participate in core services.

Cooperative Arrangements

Even Start projects are required to establish cooperative arrangements with other agencies to avoid duplicating services. This strategy allows optimal use of limited resources and allows projects to concentrate on providing new services to fill service gaps. Collaboration and cooperative arrangements were, indeed, a key focus of Even Start projects. During the 1991-92 program year, Even Start projects were involved in 2,808 cooperative arrangements to provide core services. This is an average of 27 cooperative arrangements per project. Fifty percent of the arrangements were for parenting education, 24 percent were for adult basic education, and 26 percent were for early childhood education. The most common cooperators were "other departments and programs within the public schools," "local, county, state or tribal agencies," and "postsecondary institutions."

Implementation Problems

Even Start projects reported on barriers to project implementation. The most common barriers were difficulties in the recruitment, retention, attendance, and motivation of families (41 projects), problems of communication and coordination with cooperating agencies (28 projects), financial problems (20 projects), staffing problems (16 projects), problems with facilities and space (16 projects), and problems with the evaluation (15 projects).

Participation in Even Start

• Number of Participating Families. Participation in Even Start has been increasing over time. This is due both to the addition of new projects and to improved efficiency among existing projects. The number of families served by the 76 projects first funded in the fall of 1989 increased from 2,460 in 1989-90 to 4,790 in 1990-91 and again to 5,567 in 1991-92. Thus, the same projects, with the same level of resources served more than twice as many families in their third year of operations as in their first year.

  This large increase over time in the number of families served can be attributed to the projects' ability to solve implementation problems that occur during the first year of program operations, for example, time had to be spent defining the program, recruiting staff, and setting up operations. Once these problems were solved projects became more efficient, with the extra time and resources being devoted to recruiting and serving additional families.

• Length of Participation. Although Even Start projects are funded for four years, very few families take part for that amount of time. Of the families that began Even Start in 1989-90, 55 percent participated only in that first year, 25 percent participated in both the first and second program years, and 20 percent participated in three years. This pattern improved for families in projects that were first funded in 1990, where 28 percent of the families participated for one year, and 72 percent continued into a second year.

• Reasons for Exiting. There are many reasons for turnover of families in Even Start; some are indicators of program success, while others signal problem areas. Based on families that reported a reason for leaving, completion of the planned educational program was listed by 27 percent of families that exited Even Start. Moving out of Even Start's catchment area was the most common reason for leaving the program, listed by 31 percent of families that exited. Fourteen percent of the families left because of a general lack of interest in the program. Another 14 percent reported that a family crisis of one sort or another prevented them from participating. Seven percent left the program because they became ineligible due to change in the family situation, i.e., because of federal or local eligibility requirements, there were no eligible participants in the family. Six percent gave a variety of reasons which suggested personal or structural conflicts or barriers to continued participation. These included medical reasons, work conflicts, pregnancy, scheduling conflicts, child care problems, and a lack of transportation.

• Participation in Core Services. The Department of Education requires that all Even Start families participate in each of the three core service areas during their time in the program. Almost all families had a child that participated in early childhood education during each year of the study: 90 percent in 1989-90, 97 percent in 1990-91, and 98 percent in 1991-92. Participation rates for parenting education also were quite high: 88 percent in 1989-90, 94 percent in 1990-91, and 93 percent in 1991-92. At the beginning of Even Start, participation rates were low for adult education; 54 percent in 1989-90. The Department of Education and local projects worked hard to increase these rates to 79 percent in 1990-91 and to 90 percent in 1991-92. The percentage of families participating in all three core services increased steadily over the three years of study, from 46 percent to 74 percent to 84 percent.

  It should be noted that one reason why families do not participate in all core services is that they may drop out after starting one or two services. Long-term families have much higher participation rates.

• Amount of Core Services Received. The typical family received core services in an average of between six and seven different months during their period of enrollment in Even Start. The median amount of service received by the typical family is 39 hours for adult education (mean of 86 hours), 27 hours for parenting education (mean of 49 hours), and 84 hours for early childhood education (mean of 189 hours). There is great variation in these numbers across projects.

  On a monthly basis, the typical Even Start family received 8 hours of adult education (mean of 13 hours), 4 hours of parenting education (mean of 7 hours), and 14 hours of early childhood education (mean of 25 hours). There is a clear relationship between amount of early childhood education and age of child. The typical Even Start child less than one year of age received 4 hours per month. This amount increased to 5 hours per month for one- year-olds, 9 hours per month for two-year-olds, 14 hours per month for three-year-olds, and 22 hours per month for four-year-olds.

Effects of Even Start (Short-Term)

Implementation of an Even Start project should lead to an increase in the amount of services received by families in each core service area. Appropriate support services also should be provided. While information on the types and quantities of core services provided to Even Start participants was discussed above, a key issue for attributing any observed pretest-posttest gains to Even Start as opposed to other factors is the extent to which Even Start has been able to increase families' participation in the three core service areas over what it would have been in the absence of Even Start.

Participation

Data from this study show that Even Start has substantially increased participation rates in the three core service areas. It is estimated that without Even Start, 30 to 40 percent of the Even Start population would take part in adult education. This is substantially less than the 90 percent participation rate achieved for Even Start families during the 1991-92 program year. For parenting education, participation increases from eight percent without Even Start to 93 percent for families in Even Start. And for early childhood education, participation rates are estimated to be 40 to 60 percent in the absence of Even Start and 98 percent for families in Even Start. These increases in service levels give confidence in attributing observed gains to Even Start rather than to extraneous factors such as normal maturation.

Outcomes for Children, Adults, and Families

Data to assess the short-term effects of Even Start are available from two sources: the In-Depth Study in 10 sites, which involved random assignment of families to participate in Even Start or in a control group, and the NEIS, which provides national-level data on participating families in all the projects studied. Short-term effects are measured in four areas: children, parenting literacy, parenting skills, and families.

This study has found significant positive outcomes for both children and adults. However, this is not true for all of the measures. In the absence of additional data, the impact of some Even Start services is not clear.

Effects on Children: School Readiness and Literacy

• The PreSchool Inventory (PSI). The PSI assesses a range of school readiness skills, e.g., identifying shapes and colors, and understanding numerical concepts. Children's scores on the PSI improve with age, simply as a function of maturation. This study found that children's PSI scores increase by .42 items per month due to normal maturation/development. However, by participating in Even Start, children's PSI scores increase at more than double the expected rate, by .92 items per month. This finding is based on data from the NEIS and is confirmed by data from the In-Depth Study which show that Even Start children gain significantly more than control group children on the PSI.

  This accelerated rate of learning on the PSI means that as Even Start children enter the public schools they are more likely to know basic concepts and precursors of kindergarten skills than they would have in the absence of the program. Further, the gains are similar in magnitude to those achieved by other high quality, experimental early childhood education programs.

• The Peabody Picture Vocabulary Test (PPVT). The PPVT measures receptive (hearing) vocabulary, and gives a quick estimate of verbal or literacy-related skills. Unlike the PSI, the PPVT has national norms. When administered the PPVT as a pretest, prior to entry into the program, Even Start children averaged 82 standard score points. This corresponds to the twelfth percentile nationally and highlights the low verbal skills of children at entry to Even Start.

  Unlike the PSI, the In-Depth Study and the NEIS tell different stories for the PPVT. Findings from the In-Depth Study show no significant program effect. On the other hand, data from the NEIS indicate that Even Start children gain more than would be expected without the program (0.83 standard score points per month while in Even Start vs. an expected gain of 0.0 points per month). It is possible that Even Start is having a positive effect on the PPVT, that we are seeing that effect on the NEIS, but that we are not able to detect the effect in the In-Depth Study because of a combination of small sample size and certain population effects. Additional data which may lead to firmer conclusions will be presented in the next report from this evaluation.

Effects on Parent Literacy: Literacy Gains and GED Attainment

• Literacy--the Comprehensive Adult Student Assessment System (CASAS). The CASAS is an adult-oriented functional assessment system that measures a broad range of adult literacy skills and their application in real life domains including consumer economics, government and law, occupational knowledge, community resources, and health. This evaluation used the CASAS reading survey as its measure of literacy.

  Even Start projects may have been effective in improving the functional literacy of participating adults, but results are not conclusive to date. Data from the NEIS show that adults who participate in Even Start achieve small but positive gains on the CASAS reading survey (3.6 scale score points)--gains which are comparable in size to those observed in other studies of adult education programs. However, data from the In-Depth Study show that the gains of Even Start adults are not significantly greater than the gains achieved by a randomly-assigned control group.

• Both the NEIS and the In-Depth Study show that the amount of gain on the CASAS is directly related to the amount of instruction received through Even Start, and this finding strengthens our confidence that the observed gains can be attributed to the increase in adult education participation achieved by Even Start projects. The keys to enhancing program effects in this area may be to strive for the fullest possible participation in adult education and for an increase in the number of instructional hours per month taken by participants.

• Attainment of a GED or high school diploma. Even Start has a clear, positive effect on GED attainment, an effect which is supported by data from both the In-Depth Study and the NEIS. The NEIS data on all projects show that 7.1 percent of all adults who entered Even Start without a GED or diploma achieved one while participating for a year or less in adult education services. The In-Depth Study supports the NEIS -- it shows that 14.3 percent of participating adults in the in-depth study projects' Even Start families attained a GED compared to 3.6 percent of participating adults in control group families.

  Compared to the total number of adults served, those that do attain a GED represent a relatively small percentage of the Even Start population. This may relate to the problems that are seen in retention and active participation in Even Start. Many participants do not stay with the program long enough to get a diploma or GED. Some of the adults who had not attained a GED at the end of the reporting year for this analysis remain in Even Start and may yet achieve this goal. Data on this topic will be available for the evaluation's next report.

Effects on Parenting Skills: Personal Skills, Materials in the Home, Reading Techniques, and Parents' Expectations

The study collected a variety of data on parenting and personal skills, few of which showed much change during the program. There are serious problems with some of the measures-- problems shared with other evaluations of parenting programs. This may be an area where work is needed on the evaluation measures as much as the program. Not all of the measures are problematic, however. For example, the depression measure, which shows high levels of depression in Even Start parents, may be a clear signal for greater collaboration with health services and mental health programs.

• Personal Skills. Personal skills were assessed only in the In-Depth Study. One measure of parents' personal skills is the CES-D scale of depressive symptoms. Scores above a certain level on the CES-D are an indication of clinical depression. Based on this test, 45 percent of Even Start parents in the In-Depth Study sites were clinically depressed at entry. This result is similar to those reported in other studies of low-income populations. At the posttest, no effects were detected on the incidence of depressive symptoms. A second measure of parents' personal skills is Pearlin and Schooler's Mastery Scale, a measure of self-efficacy or locus of control. As with depressive symptoms, no effects of Even Start on parents' self-efficacy were detected.

• Home Learning Environment. Several scales describing different aspects of the home learning environment were included both in the In-Depth Study and in the NEIS. The In-Depth Study found positive program effects only on one of these scales: reading materials in the home. This scale measures the number of different types of reading materials found in the home, e.g., books, magazines, newspapers. Although Even Start families showed positive gains on another scale (amount of play materials in the home), these gains were not significant given changes in the control group. No change was observed for several other scales including measures of learning activities in the home, teaching child at home, talking with child at home, family rules, activities outside home, and the parent as a teacher scale.

  On the other hand, data from the NEIS survey of all Even Start families show small but statistically significant gains on each measure of the home learning environment. The size of the gains are roughly what was seen for the In-Depth Study, however the NEIS data are based on a much larger sample which allows greater statistical power. Without a control group it is difficult to tell whether the observed gains are attributable to Even Start, and so the effects of Even Start on the home learning environment are mixed: we see clear, positive effects only on the amount of reading materials in the home, but there are indications that gains may be occurring in other areas.

• Parent-Child Reading Task. The In-Depth Study included a newly-developed measure of parent-child reading interactions. The parent was asked to read a simple book to her child, while a trained observer used a pre-coded rating form to record five major aspects of parent-child interactions: parent describes book to child, parent questions or responds to child, child responds, parent's general affect, and child's general affect.

  There was no effect of Even Start on any of the five areas of parent-child reading observation. While children did show a significant increase in their involvement and interaction during the reading activity (child's general affect) this increase was not significant over and above the change seen for the control group.

• Parental Expectations. In both the In-Depth Study and the NEIS, parents were asked questions about how well they expected their child to do in high school and the likelihood that their child will graduate from high school. As is the case for such questions in most studies, parents generally had high expectations for their children at the pretest. Even so, Even Start significantly increased parental expectations as assessed in both the In-Depth Study and the NEIS This is a heartening finding--parental expectations are an important factor in how well children do in school.

Effects on Families: Social Support, Income, and Employment

The In-Depth Study included several measures of the effects of Even Start on participating families, including their perceived social support, adequacy of financial resources, income level, and sources and employment status. The NEIS instruments had questions about income and employment. Across these measures, gains from program entry to the end of the first program year were minimal. It is possible that these areas are difficult to change because they are affected by circumstances beyond the control of the program, such as the local economy and the availability of jobs. It is also likely that these indices will not show large changes in the short-run, but instead require longer interventions and other positive short-term impacts in order to achieve significant gains.

While there are few effects for families on these quantifiable variables, Even Start participants describe a number of qualitative changes in their lives and the lives of their children. Listening to the personal stories of program participants, it becomes apparent that most of the changes in attitudes and skills that the parents see in themselves and their children are positive short-term impacts of the program. Parents describe themselves as moving toward their goals of an educational certificate, getting a job and being a better parent. Given the current status of Even Start parents, it is reasonable that we do not yet see changes in the more distant outcomes of increased employment and income.

Final Comments on Outcomes

Even with the lack of power in the In-Depth Study and the differences between the small, more powerful In-Depth Study and the NEIS study of all sites, the overall picture is encouraging. Data from both the NEIS and the In-Depth Study show positive effects for children and for parents. There are no positive effects identified for families, but the outcomes (e.g., income, employment) in this area tend to be ones that Even Start does not expect to change in the short term. It will be important to determine whether the positive short-term effects for children and parents deteriorate over time or whether they are maintained, and whether it is possible to detect long-term effects for families.

Cost of Even Start

The federal cost per Even Start family has declined over the life of the program, from $5,894 in 1989-90 to $3,669 in 1990-91, and again to $2,503 in 1991-92. This is due to increases in the number of families served each year, suggesting that over time, projects have matured and become more efficient.

Even Start projects also obtain substantial resources (e.g., matching funds, in-kind contributions, and the value of services provided by cooperating agencies), in addition to their federal Even Start funds, in order to deliver appropriate services to participating families. In 1991-92, $2,503 in federal funding per family was augmented by $1,352 in other resources to arrive at total resources of $3,855 per Even Start family. Thus, federal Even Start funds average 65 percent of the total resources per family and other funds average 35 percent.

Almost two-thirds (64 percent) of all Even Start resources were spent on the direct provision of services: 31 percent for early childhood education, 15 percent for adult education, 9 percent for parenting education, and 9 percent for support services. Even Start funds also were spent for program administration and coordination (14 percent), evaluation (10 percent), case management and recruiting (4 percent), and for a variety of other functions (8 percent). Many Even Start programs are building on existing resources to a significant degree, as was intended by Congress.

Even Start costs vary tremendously across projects. As might be expected, projects which serve large numbers of families do so at a lower cost per family. In particular, projects which serve 100 or more families do so at a federal cost of $1,659 per family, while projects which serve 30 or fewer families spend an average of $6,312 per family. Also, projects which coordinate with other agencies to provide core services have a lower cost ($1,878 per family) than projects which retain primary responsibility for providing core services (cost of $5,775 per family).

Copies of the full report, National Evaluation of the Even Start Family Literacy Program: Report on Effectiveness may be obtained from:

U.S. Department of Education
Office of the Undersecretary
Planning and Evaluation Service
600 Independence Ave, SW
Room 4163 FOB-10
Washington, DC 20202-8242

202 401-3630

-###-

Return to Elementary and Secondary Education Page

mail to esed@ed.gov

Last update September 1996 (swz).