Head Start Family and Child Experiences Survey (FACES): 1997 Cohort [United States] (ICPSR 4134)
Published: Jun 30, 2014
The Head Start Family and Child Experiences Survey (FACES) is an ongoing national longitudinal study of the cognitive, social, emotional, and physical development of Head Start children. It examines the characteristics, well-being, and accomplishments of families, the observed quality of Head Start classrooms, and the characteristics and opinions of Head Start teachers and other program staff. FACES was designed to address four central questions related to program performance objectives:
- Does Head Start enhance children's development and school readiness?
- Does Head Start strengthen families as the primary nurturers of their children?
- Does Head Start provide children with high quality educational, health, and nutritional services?
- How is classroom quality related to child outcomes?
The FACES 1997 Cohort involved a nationally representative sample of children and families in Head Start programs in the United States who were studied at entry into the program in the fall of 1997, assessed in the spring at the completion of one or two years of Head Start, and followed up in the spring of the kindergarten and first grade years. During these visits, the research team completed individual interviews with staff and parents, child and classroom observations, direct child assessments through one-on-one interviews, and indirect assessments of children by their teachers and parents.
The data collection instruments used in FACES were designed to measure several different types of skills, accomplishments, and behaviors that are relevant to a child's school readiness and social competence:
- The parent interview was designed to collect up-to-date information about current Head Start families while being sensitive to differences based on the background of the respondents. The interview provided descriptive information about the parents (education, work status, health, nativity, depression, social support, use of discipline and rules, exposure to violence), the household (income, housing, activities with children, use of child care), and the children (gender, ethnicity, health, behavior, literacy skills, disabilities, exposure to violence). Additionally, parents reported how their families came to Head Start and how they perceived their Head Start experiences.
- The staff interview was designed to provide a profile of the background, qualifications, and training of Head start personnel as well as an understanding of classroom activities, family activities, services, local programs offered to families, and staff perspectives on their programs and the families they served.
- The child and classroom observations were designed to record information from the observations of children's behavior and home environments. Some of the assessments used included Peabody Picture Vocabulary Tests (PPVTs), Woodcock-Johnson Revised Tests of Achievement, Norm Referenced Cognitive Tests, Story and Print Concepts, and McCarthy Scales of Children's Abilities.
United States Department of Health and Human Services. Administration for Children and Families. Office of Planning, Research and Evaluation (Contract number HHS-105-96-1912, Contract number HHS-105-96-1930 )
This data collection may not be used for any purpose other than statistical reporting and analysis. Use of these data to learn the identity of any person or establishment is prohibited. To protect respondent privacy, the FACES 1997 Cohort data are restricted from general dissemination. Access to parts of this study requires a signed User Agreement. To obtain the file(s), researchers must agree to the terms and conditions of the Restricted Data Use Agreement, which is included with every download and can also be obtained separately on the Browse Documentation page.
1997 -- 2001 (Fall 1997 through Spring 2001)
Date of Collection
1997-09 -- 2001-06
Data Collection Notes
The Head Start Family and Child Experiences Survey would like to acknowledge Louisa Tarullo's role as the Project Officer.
Nicholas Zill, Ph.D. (Westat) was the Principal Investigator for this project.
Special collaborators for the collection and preparation of the data files that comprise the FACES dataset include Ruth Hubbell McKey, Ph.D. (Xtria), David Connell, Ph.D. (Abt Associates, Inc.), Robert O'Brien, Ph.D. (CDM Group, Inc.).
Several reports based on this data collection are available at the Administration for Children and Families Web site.
To conform with ICPSR filenaming conventions, data and documentation files have been assigned new file names that differ from those assigned by the original data producer, Westat, Inc. A crosswalk document (linking old data file names to new ones) may be found within each of the codebooks and the User Guide.
A representative sample of 3,200 children who were three and four years old and their parents in a stratified national probability sample of 40 Head Start programs. The sample was stratified by three variables: region of the country, urbanicity, and percentage of minority families in the program. Migrant and Seasonal Head Start programs and American Indian programs were excluded from the sampling frame.
The programs participating in the FACES 1997 Cohort were randomly selected from 1,734 Head Start programs that operated during the 1995-1996 program year in the 50 United States, Puerto Rico, and the territories of the United States. The available Head Start programs served approximately 785,000 ethnically diverse children aged three and older.
Mode of Data Collection
Original Release Date
2014-06-30 Internal records have been edited.
2013-07-08 The FACES instrument matrix has been updated to include the 2009 cohort.
2013-05-24 The FACES instrument matrix has been updated and the FACES-produced questionnaires are now available in the documentation download.
2010-12-16 The Restricted Data Use Agreement has been updated.
2010-10-27 Updated the FACES Instrument Matrix to include the 2006 cohort.
2010-06-30 The FACES Instrument Matrix and public PDF versions of the Codebooks have been added.
2010-02-05 The Restricted Data Use Agreement was updated.
2008-09-07 The Restricted Data Use Agreement has been added.
The FACES 1997 data include five sets of weights. Cross-sectional weights were generated for the fall 1997 and spring 1998 data, with additional weights created for use with the longitudinal findings. Each set of weights includes a full-sample weight and replicate weights. There is also a single set of jackknife factors for use with all of the weights.
Using program-level information from Head Start's Program Information Report (PIR) and center-level information collected directly from the programs, three levels of weights -- program, center, and child -- were generated. For each child, the final child weight = (program weight) x (center weight) x (child weight).
The fall 1997 child cross-sectional weights were calculated as the inverse of the product of the probabilities of selection at each stage of sampling. A final adjustment was made to each of the child weights so that they represented the full population of Head Start children. This adjustment was made by multiplying each child weight by the ratio of the expected number of children in Head Start in each category to the sum of the weights of the actual children in the study. As a result of the weighting procedure, the fall 1997 sample was weighted to represent a Head Start population of 779,785.
The three spring 1998 child cross-sectional weights were generated by making adjustments to the original fall 1997 cross-sectional weights to account for the change in sample size from fall to spring. As a result of the weighting procedure, the spring 1998 sample was weighted to represent a Head Start population of 763,671.
The child longitudinal weights were generated for two groups of families: (1) those families in which the same respondent participated in both the fall 1997 and the spring 1998 parent interviews, and (2) those families in which the same respondent participated in the fall 1997, the spring 1998, and the spring 1999 parent interviews. In each case, the fall 1997 child weight was adjusted for nonresponse by multiplying the weight by a program-level factor that accounted for the number of families that had different interview respondents over time or who did not complete the interview due to refusal, an inability to contact the family at the time of the visit (although the family was still enrolled in Head Start), or the inability of the parent to be available to the interviewers during the time of the site visit. The application of this weighting procedure for the longitudinal sample (children who were in Head Start from fall to spring) resulted in a representation of 634,949 children.
Jackknife Replicate Weights
For each of the 40 Head Start programs, a set of 40 jackknife replicate weights was created for calculating standard errors. The replicate weights were created using a stratified jackknife procedure. One program at a time was dropped and the weights of the remaining programs in the same stratum were adjusted by a factor of nh/(nh-1), where nh is the number of sampled programs in stratum h. The program weights in the other strata were left unchanged. By repeating this 40 times, 40 replicate weights were obtained for each program.
Consult the Data User's Guide for a more in-depth explanation of the weighting procedure and the specific formulas that were used for each of the weights.
The public-use data files in this collection are available for access by the general public. Access does not require affiliation with an ICPSR member institution.
This study is provided by Child Care & Early Education Research Connections.