Table of Contents | Previous | Next |
NATIONAL LONGITUDINAL STUDY OF ADOLESCENT HEALTH (ADD HEALTH)
Purpose
The National Longitudinal Study of Adolescent Health (Add Health) determines the factors that contribute to the physical and mental well-being of adolescents. The survey covers multiple contexts, including family, friends, peers, school, neighborhood, and community.
Agencies/Institutions
The primary funding of Add Health was from the National Institute of Child Health and Human Development (NICHD). Seventeen other Federal agencies also provided some funding. Quality Education Data, Inc. (QED) provided the database used to generate the random sample of U.S. schools.
The National Opinion Research Center (NORC) of the University of Chicago fielded Waves I and II of the study; the Research Triangle Institute (RTI) conducted the fieldwork for Wave III.
Research/Survey Design
Add Health is a nonexperimental, nationally representative, longitudinal study of students in grades 7 through 12 in the United States in the 1994–1995 school year. Data were collected from the youth, their parents, siblings, friends, romantic partners, fellow students, and school administrators through multiple data collection components, including an adolescent in-school survey, an adolescent in-home interview, a parent in-home interview, and a school administrator survey. All instruments were fielded in Wave I. Wave II included an adolescent in-home interview as well as telephone updates from the school administrator. Wave III consisted only of a respondent in-home interview. Available data also include picture vocabulary test scores, an in-school friendship network dataset, and information on the geographic location of households within the communities.
The study used a multistage, cluster sampling design in which the school was the primary sampling unit (PSU). U.S. high schools were stratified into 80 clusters by school characteristics and were randomly sampled from schools in a QED database. The study design required a school to have students enrolled in grade 11 and a total population of at least 30 students. The stratification characteristics and categories can be found below:
- Region: Northeast, Midwest, South, West
- Urbanicity: urban, suburban, rural
- School size: 125 or fewer, 126–350, 351–775, 776 or more students
- School type: public, private, parochial
- Percent white: 0, 1 to 66, 67 to 93, 94 to 100
- Percent black: 0, 1 to 6, 7 to 33, 34 to 100
- Grade span of the school: K to 12, 7 to 12, 9 to 12, 10 to 12, vocational/technical, alternative, special education
If the selected high school did not include all targeted grades, its feeder middle school was also eligible for the study as long as the school had grade 7 students enrolled. In total, 80 high schools and 52 feeder middle schools were selected.
The in-school survey was administered to all students in the selected schools, provided they agreed to participate and were not absent the day the survey was conducted. A nationally representative core sample of 12,105 students who completed the initial in-school survey was targeted for follow-up in the in-depth Wave I in-home interviews. To determine the individual students who would make up the Wave I core sample cohort, approximately 200 students from each school pair, stratified by grade and sex, were targeted.
In addition to the core sample, the study also oversampled students from several special subgroups: disabled youth; Chinese, Cuban, and Puerto Rican adolescents; and black youth from high socioeconomic status families. Families were considered to be of high socioeconomic status if at least one parent held a college degree. In addition, the study oversampled adolescents living together within one household. This group is referred to as the “genetic sample” and includes oversamples of twins, half- and step-siblings, and nonrelated pairs. Some of the adolescents selected from this oversampled group did not attend one of the original 80 high schools or 52 middle schools but were recruited because they resided with an adolescent who did attend one of those 132 schools.
Oversampling occurred at the school level as well. Of the 132 schools, 2 large schools (enrollment greater than 3,300 students) and 16 smaller schools (enrollment fewer than 300 students) were selected as the “saturation sample” in which all students enrolled in these schools were asked to participate in the in-depth in-home interviews. Collecting detailed information from all students in the saturation sample allows a more thorough analysis of adolescents’ social networks.
Date(s)/Periodicity
Wave I of the study was fielded between September 1994 and December 1995. Wave II of the study was fielded approximately 1 year later (1996), with Wave III following approximately 6 years later (2001–2002).
Population/Sample
More than 70% of the originally targeted high schools agreed to participate in the study. A replacement school was selected from the same stratum if a school declined to participate.
From school recruitment, 90,118 students responded to the in-school questionnaire. A subset of 20,745 respondents then completed the Wave I in-home questionnaire. This “nationally representative sample” from Wave I included a core sample of 12,105 teens and all the oversampled groups except the genetic sample. The full Wave I sample was weighted to be nationally representative of adolescents grades 7 through 12 in the United States in 1994–1995.
The supplemental sample of students with disabilities comprised about 2.8% of the total Wave I sample, and Chinese, Cuban, Puerto Rican, and high socioeconomic black youth (the ethnic oversample) comprised approximately 10.9% of this sample. Mexican-American, Nicaraguan, Japanese, South Korean, Filipino, and Vietnamese youth were also represented in the core sample.
In 1996, 14,738 students (71% of the Wave I respondents) completed the Wave II in-home questionnaire. The Wave II survey did not follow up on adolescents who were in grade 12 at Wave I. Thus, when only eligible respondents are considered, the Wave II response rate was 88.2%.
Wave III reinterviewed 15,197 of the original Wave I respondents (now ages 18–26) as they entered young adulthood. This survey excluded original participants who could not be located and those who were located but were living overseas or were in the military. The Wave III sample did include respondents who were incarcerated. In addition, the Wave III survey collected information from 1,507 relationship partners of the original respondents. These additional respondents were divided evenly among married, cohabiting, and dating partners.
Content Covered
Topics in the adolescent in-home questionnaire were daily activities, health status and access to health care, educational experiences, relationships with family and peers, romantic relationships, nonromantic sexual relationships, motivations for birth control, physical development, use of illicit substances, delinquency, joint occurrences of delinquent behavior with the use of illicit substances, fighting and violence, expectations, protective factors, employment, emotions, self-efficacy, suicidality, religion, and neighborhood characteristics. Questions were also asked about the youth’s knowledge about sexual intercourse and contraception and perceptions about pregnancy, AIDS, and STD risk.
Parents were asked questions on their employment, income, health, marital history, and relationship with their current partner. In addition, parents were asked child-specific questions about the adolescents’ social life (romantic and nonromantic), friends, and health experiences and conditions.
In Wave III, questions were asked about relationship, childbearing, educational, and employment histories, and the picture vocabulary test was readministered. Also, respondents were asked to provide saliva and urine specimens so that the prevalence of HIV and other sexually transmitted diseases among young adults could be studied.
Availability of Data for Public Use
The official Add Health website is http://www.cpc.unc.edu/addhealth.
Select data from one-half of the core sample and one-half of the high socioeconomic black student sample are available in a public-use dataset. Selected community contextual variables are also available.
An application and agreement for data use is available at http://www.socio.com/agreepay.pdf.
For more information:
Sociometrics Corporation
170 State Street, Suite 260
Los Altos, CA 94022
1-800-846-3475
socio@socio.com
Reference List for Users’ Guide, Codebooks, Methodology Reports
Carolina Population Center. (2000). Research design: Facts at a glance. Chapel Hill: University of North Carolina.
Carolina Population Center. (2002). Add Health funding. Chapel Hill: University of North Carolina.
Carolina Population Center. (2002). Research design: The Study. Chapel Hill: University of North Carolina.
Carolina Population Center. (2003). Add Health research design. Chapel Hill: University of North Carolina.
Carolina Population Center. (2003). Design focus. Chapel Hill: University of North Carolina.
| Table of Contents | Previous | Next |