ACF OPRE: Head Start Family and Child Experiences Survey (FACES)

Executive Summary

In 1997, Head Start launched the Family and Child Experiences Survey (FACES), a study with a nationally representative sample of Head Start programs to describe the characteristics, experiences, and outcomes for children and families served by Head Start. It also explores the relationships among family and program characteristics and outcomes. In 1997, the FACES design included a nationally representative sample of 3,200 3- and 4-year old children and their families in 40 programs. The FACES 2000 sample includes 2,800 children and their families in 43 different Head Start programs across the nation.

Each cohort of FACES employs a nationally stratified sample of Head Start programs, centers, classrooms, children, and parents. FACES 2000 features four phases of data collection and follows 3- and 4-year-old children from program entry, through one or two years of program experience, with followup in the Spring of kindergarten. The FACES 2000 battery has four main components: the direct child assessment, parent interview, teacher and staff interviews, and classroom observations. Although there is no non-Head Start comparison group in FACES, the use of assessment measures with national norms permits comparisons between the skills of children in the sample and children of the same ages in the norming samples.

ES.1 Study Findings

FACES provides information about the knowledge and skills that children have when they enter the Head Start program and the gains they make during the Head Start year and the first year of elementary school. It also describes the quality of Head Start classrooms, and factors that help explain variations in quality across Head Start classrooms. In addition, FACES 2000 data provide insight into the relationship of program and classroom characteristics to children’s outcomes, as well as the relationship of family and parental characteristics to children’s outcomes.

ES.2 Who is Head Start Serving?

According to data from FACES 2000, Head Start is serving a diverse population of children. Slightly more than one-third (35 percent) of Head Start children were White followed closely by African American (32 percent) and Hispanic (28 percent). The majority of children (62 percent) were four and five years of age and 38 percent of the children were three years of age. Slightly more than two-fifths (41 percent) of the families were located in the southern region of the country and the vast majority of the families (70 percent) were situated in urban/suburban areas.
Head Start families face numerous risks and challenges that are related to children’s well being. Parents (mother or father) of Head Start children had a high school diploma or GED (41 percent), had a vocational/technical diploma or an Associate’s degree (30 percent), or had a Bachelor’s degree or higher (6 percent). Parents reported working on a full-time basis (39 percent) or on a part-time basis (14 percent), and 34 percent of parents were not in the labor force.
The majority of Head Start families reported monthly family incomes of either $500 to $999 (27 percent) or $1,000 to $1,499 (25 percent). Nearly one-half (48 percent) of the children resided in dual-parent families whereas more than one-half (52 percent) of children resided in other family arrangements. One-half (50 percent) of the Head Start families were comprised of 4 to 5 persons.

ES.3 Head Start Children’s Cognitive and Social-Emotional Outcomes

Cognitive assessment data from FACES 2000 showed that most children entered Head Start with early literacy and math skills well below national averages. However, there was considerable diversity in skill levels among Head Start children. At Head Start entry, the highest quarter of Head Start children were at or above the national average (50^th percentile) in early language and number skills, while the lowest quarter of children ranked in the lowest 2 percent of all U.S. preschoolers in these areas.
Children made gains toward national averages during the Head Start year, especially with respect to vocabulary knowledge and early writing skills. Children who entered the program with lower levels of knowledge and skill showed larger gains during the program year, yet still lagged considerably behind national averages. Children who started with higher assessment scores in the Fall wound up with higher scores in the Spring, but did not show the same level of gains as those with lower baseline scores.
Language-minority children in Head Start showed significant gains in English vocabulary skills without declines in their Spanish vocabulary skills. They did not gain in letter recognition skills.
Children who graduated after two years of Head Start made greater gains from entry to graduation than those who spent only one year in Head Start. Typically, children who graduate Head Start after 2 years enter with lower skills than one-year graduates. However, by the time both groups leave Head Start, they achieve similar levels of performance.
Head Start graduates also showed further progress toward national averages during kindergarten. Gains of between one-quarter to more than three-quarters of a standard deviation were observed in vocabulary, early math, and early writing skills during kindergarten. Most Head Start graduates could identify the letters of the alphabet by the end of kindergarten and more than half could recognize beginning sounds of words. Further, Head Start graduates achieved early writing scores near the national norm by the end of kindergarten. Nevertheless, Head Start graduates remained behind their more advantaged peers in vocabulary and early math.
FACES 2000 findings indicate that teachers reported significant positive changes on measures of social competence from Head Start entry to graduation.
The findings also indicate that children who spent 2 years in Head Start were reported by teachers to display higher levels of cooperative classroom behavior and social skills and less hyperactive behavior at the end of Head Start compared to children who spent only 1 year in Head Start.
Children entering Head Start with the lowest levels of social skills, and those who entered Head Start with the highest levels of problem behaviors, gained the most during their participation in Head Start. Although these children did not close the gap between themselves and other children, they did narrow it significantly.
At the beginning of the Head Start year children’s disability status was highly associated with both social-emotional and cognitive developmental outcomes. On average, children with disabilities did not perform as well as other children on all baseline social-emotional and cognitive outcomes with the exception of book knowledge. White children were more likely than African American children to have a diagnosed disability, and boys were more likely than girls to have a disability, consistent with findings from other studies. Speech and language impairments constituted the largest group within the disabilities categorization.
In addition, children with disabilities did not progress as well as other children on more than half (9 of 16) of the cognitive and social-emotional outcomes, including parent- and teacher-reported hyperactivity, withdrawn behavior, total problem behaviors, early math skills, perceptual–motor skills, and book knowledge. There were no statistically significant differences on the remaining seven outcomes between the two groups.

ES.4 Head Start Program’s Use of Curricula

Findings from FACES 2000 indicate that the great majority of Head Start programs use a curriculum, with the goal of providing a planned, developmentally appropriate early childhood program for children. Most teachers reported using a single curriculum. The most frequently used curricula are the Creative Curriculum (39.1 percent) and High/Scope curriculum (20 percent). Further, 93 percent of teachers reported receiving training in their curriculum, and 92 percent reported receiving ongoing support.
Data from FACES 2000 also indicate a relationship between program characteristics and the type of curriculum used. Specifically, classrooms using a curriculum other than the Creative Curriculum or High/Scope curriculum served the poorest Head Start families and the highest percentage of non-White children. There was also regional variation in curricula used by classrooms. More teachers from the Northeast (78.7 percent), Midwest (59.4 percent), and West (69.4 percent) used either the Creative Curriculum or High/Scope curriculum compared to any other type of curricula, while the majority of teachers from the South (54.4 percent) used a curriculum other than the Creative Curriculum or High/Scope curriculum.
Findings also demonstrate a relationship between type of curriculum and classroom quality. Classrooms using Creative Curriculum or High/Scope had significantly higher scores on three measures of classroom quality (the Early Childhood Environment Rating Scales-Revised—ECERS-R—Total score, ECERS-R Language subscale score, and a Quality Composite score) than classrooms using curricula categorized as “Other.” Analyses were also conducted separating the “Other” curricula category into Widely Available Curricula and All Other Curricula. Classrooms using Creative Curriculum, High/Scope curriculum, and Widely Available Curricula had similar average ECERS-R Total scores and ECERS-R Language subscale scores, with classrooms using All Other Curricula receiving lower average scores on these instruments. However, on the Quality Composite measure, classrooms using Creative Curriculum or High/Scope had similar scores, with classrooms using either Widely Available Curricula or All Other Curricula receiving lower marks. Multivariate analyses also showed significant differences between classrooms using Creative Curriculum or High/Scope compared to All Other Curricula on all three classroom quality scores.

ES.5 Quality in Head Start Classrooms

Head Start classrooms continue to show good levels of quality, based on the indicators of quality measured in FACES 2000 in the Fall 2000, Spring 2001, and Spring 2002. These levels of quality are consistent from FACES 1997, and the consistency is evident across a wide variety of the indicators, including child:adult ratio, teacher-child interactions, and classroom activities and materials. In fact, FACES shows that Head Start has a better, more limited range of quality than that seen in child care centers and preschools in several other national studies, with few Head Start classrooms scoring below minimal quality.
Head Start teachers are qualified and experienced (although as a group they do not have the same level of credentials as public school teachers), and there appear to be substantially more teachers with higher educational attainment in this cohort compared with the 1997 FACES cohort. Head Start teachers in 2000 are younger, compared with those in 1997-1998, and more of them have been teaching in Head Start for two years or less. These newer teachers are also the ones most likely to have a graduate school degree.
Classrooms with higher levels of quality are those whose teachers have higher levels of education, experience, and knowledge and attitudes of early childhood education practices. The relationship between teacher education and classroom quality is explained by teacher’s attitudes and knowledge of early childhood education practices. Teachers who are more educated have more positive attitudes and knowledge, which translates into higher levels of classroom quality.
Programs using one of the two most widely used integrated early childhood curricula (i.e., High/Scope and Creative Curriculum) also have teachers with positive attitudes and knowledge about early childhood education practice. Both of these factors appear to have the strongest effect on teacher sensitivity and responsiveness compared with other indicators of quality.
Multilevel analyses suggest that variations in the quality of Head Start classrooms may be explained by characteristics of the families and children they serve, by the curriculum used in the program, and by teacher attitudes and knowledge about early childhood education practice. The results suggest that Head Start classroom quality may be affected by factors beyond the classroom door that are characteristics of the program and the families who participate.

ES.6 Relationship of Program and Classroom Characteristics to Children’s Outcomes

Program- and classroom-level variables provided unique and significant contributions to the prediction of children’s assessment scores at Head Start entry and graduation, as well as gains from entry to graduation. Analysis of data from FACES 2000 showed that children’s cognitive gains in Head Start were significantly related to use of the High/Scope, higher teacher salaries, teachers' educational credentials (Bachelor's degrees or Associate’s degrees or higher), and provision of preschool services for a longer period each day.
For each cognitive measure, children who attended Head Start for two years showed greater graduation scores and greater entry to graduation gains compared to their peers who attended Head Start for only one year.
Surprisingly, within the narrow range of child:staff ratios in Head Start, higher child:staff ratios were associated with greater entry-graduation achievement gains in some of the cognitive outcomes. Further, teachers’ scores on a scale assessing their knowledge of early childhood education practices were negatively linked with children’s cognitive outcomes.
The strongest predictors of children’s social behaviors were predictors at the child and family level, mainly the child’s age, gender, disability status, and number of years spent in Head Start, as well as the presence of both biological or adoptive parents in the house and the existence of books in the home. Most child and family characteristics had stronger links with the social behavior variables than program or classroom characteristics.
The program- and especially the classroom-level characteristics showed relatively low power to predict children’s social behaviors. Nonetheless, the addition of the program and classroom levels to the child- and family-level set of characteristics did result in higher explanatory power of the combined model over and above the child- and family-level model by itself. This addition was significantly higher in almost half of the variables examined. Thus, the inclusion of program- and classroom-level sets of variables in the combined models increased the models’ overall fit to the data. This finding supports the expectation that the Head Start experience may serve to compensate for child and family disadvantages.
Differences in cognitive gains between programs and classrooms possessing certain quality indicators (such as the use of High/Scope, higher teacher salaries, teachers with Bachelor's or Associate's degrees or higher, and provision of preschool services for a longer period each day) and programs and classrooms without those features, while statistically significant, were relatively modest in magnitude. By itself, each of the differences was not large enough to close the gap between where Head Start children typically end up at the end of the program year and the average achievement levels of American children at the start of elementary school.
The ECERS-R Language subscale was not associated with greater gains on any of the cognitive measures. The failure to find significant links between children’s cognitive gains in Head Start and class-level scores on the ECERS-R Language subscale may have to do with the generally good quality of Head Start classrooms and the limited range of variation in classroom quality that FACES found in its national samples of programs and classes. Studies encompassing broader ranges of quality of childcare and early education facilities have shown greater variations in classroom quality measures and significant relationships between quality measures and children’s gains.

ES.7 Changes in Head Start Families Over Time

Data from FACES 2000 indicate that the condition of Head Start children’s families and especially their mothers has generally remained the same or marginally improved as children move from Head Start to kindergarten. Although not all mothers reside in the household of their children, the percentage remains consistently over 90 percent. A high percentage of mothers are also single, never married mothers. Over 40 percent of households have only the mother (and no father) residing in the child’s household.
The longitudinal analysis shows evidence of an improvement in the socioeconomic status of households over the years. The data show that a greater percentage of households report higher levels of monthly income in Spring 2002 and Spring 2003. Moreover, a smaller percentage of households remains in poverty or depends on welfare as children move through the Head Start program and on to kindergarten.
Examining the relationship between these mothers’ and household characteristics and children’s cognitive outcomes using the FACES 2000 data shows that a change in the family situation represented in terms of a change in mothers’ marital status or coresidence with parents bears little relationship to children’s cognitive outcomes. Instead, household characteristics such as parents’ education and household poverty status are more important. While greater parents’ education results in better cognitive outcomes, the poor household economic situation is associated with lower assessment scores.

ES.8 Relationship of Family and Parental Characteristics to Children’s Outcomes

One-quarter of the parents were classified as moderately or severely depressed. Parents who were more depressed reported that their children had more problem behaviors and fewer positive social behaviors, a finding supported by the teachers’ reports of children’s behavior. Their children also had lower scores on one-to-one counting, creativity, design copying, early writing, letter identification, and early math assessments.
More than one-fifth of the parents reported they had witnessed violent crime. Five percent reported being victims of violent crime in the neighborhood, while a similar percentage reported being victims of violence in their homes. Almost 10 percent of the children were reported to have witnessed domestic violence during the previous year. Less than 2 percent of the children were reported to have been victims of violent crime or victims of domestic violence. Positive correlations were found between increased exposure to neighborhood violence and reports of child problem behavior, while children in more violent neighborhoods had lower assessment scores on the color naming and book knowledge assessments.
Almost 13 percent of the parents indicated that they have been victims of domestic violence. Teachers and parents reported children in these families had more overall problem behaviors.
Almost one-half of the Head Start children lived in households with at least one individual who smoked cigarettes and about two-fifths of the households reported having at least one individual who drank alcohol. Living in a household with someone who drinks increased the risk of parental depression, while the children in these homes were reported to have more overall problem behavior and scored lower on vocabulary, color naming, and social awareness assessments.
Almost one-fifth of the parents reported that someone in their household had been arrested and charged with a crime. Children in these families were more than 3 times more likely to have been a witness to violent crime or domestic violence in the past year. These children were also 3 times more likely to have been a victim of domestic violence or violent crime. These children had lower vocabulary scores, and were reported by both parents and teachers to be more aggressive and have more overall problem behaviors.
At least one of a set of selected risk factors was evident in over 90 percent of the families. Nearly one-quarter of the families had four or more risk factors. Children in the families with four or more risk factors had lower parent ratings on emergent literacy and higher teacher and parent ratings of problem behavior. In the assessments, these children scored lower on design copying, color naming, one-to-one counting, book knowledge, vocabulary, early math, early writing, and letter identification.
Analyses of child-level data show that Head Start children whose parents read to them more often show greater gains and graduation scores on some measures.
Families engaged their children in a number of weekly and monthly activities. The number of activities was positively correlated with positive child behaviors and emergent literacy and negatively correlated with problem behaviors. In particular, the weekly activities had positive correlations with scores on the social awareness, color naming, one-to-one counting, book knowledge, vocabulary, early math, early writing, and letter identification tasks.
Higher authoritative parenting style scores were significantly positively correlated with children’s social awareness, but not with any of the other cognitive measures. On the other hand, higher authoritarian scores were significantly negatively correlated with comprehension, color naming, vocabulary, and early math assessments.
More than two-thirds of parents had attended parent-teacher conferences, observed in their children’s classrooms for at least 30 minutes, or met with a Head Start staff member in their homes. Parent involvement in Head Start was positively correlated with parental reports of positive social behavior and higher emergent literacy skills and negatively correlated with aggressive and overall problem behavior. Children whose parents were more involved in Head Start scored higher on vocabulary, book knowledge, early writing, early math, and letter identification tasks.
Parent involvement at Head Start, parent reports that they and their children had positive experiences at Head Start, or parent satisfaction with the program significantly moderated relationships between risk factors (e.g., parental depression, exposure to violence and domestic violence, substance use, and involvement with the criminal justice system) and many negative child behavior and lower cognitive outcomes. These findings suggest that Head Start may play an important role in protecting families and children from the challenges that low-income families face.

ES.9 Predictive Validity of the FACES Cognitive and Behavioral Measures

The FACES measures have strong predictive validity with outcomes at the end of kindergarten. As an indicator of early literacy skills, the cognitive measures show strong associations with reading ability at the end of the kindergarten year. As an indicator of school adjustment and social competence, the behavior ratings demonstrate ability to predict kindergarten behaviors that promote learning. The instruments used in FACES also predict whether a child gets promoted to first grade.
The instruments used in FACES may tap different types of abilities (“inside-out” or decoding skills vs. “outside-in” or comprehension skills) that are important for children’s future reading proficiency and academic achievement.
In the assessment of children’s social competencies, the use of parent and teacher ratings provides data on children’s behavior skills in different situations and provides a comprehensive picture of their behavior. Equally important, both parent and teacher ratings significantly contribute to the prediction of social skills at the end of kindergarten. The parent and teacher ratings also significantly predict reading skills and general knowledge at the end of kindergarten. Ratings of problem behaviors were negatively correlated with kindergarten Reading and General Knowledge scale scores, suggesting that behaviors that may impede learning are associated with lower reading skills in kindergarten. High ratings of behaviors that enhance learning, positive approaches to learning and cooperative classroom behavior were positively correlated with kindergarten outcomes.

Table of Contents | Previous | Next