APPENDIX 2.1: SAMPLING DETAILS
CREATING THE IN-SCHOOL SAMPLE
Harris Interactive, Inc., who conducted the sampling and administered the survey, has developed a sampling process and survey methodology for surveying representative samples of U.S. school students. All interviewing is conducted in the classroom.
The Harris national probability samples of schools and students are based on a highly stratified two-stage sampling design. This design employs features similar to sample designs used in various national surveys of students and schools conducted by the National Center for Educational Statistics. In the first stage, a sample of schools is selected from a list of all schools. In the second stage, a class is randomly selected to participate from within the chosen school.
The school sample is drawn from a list of approximately 80,000 public, private, parochial schools in the United States. It is selected to account for differences in grade enrollment, region, and the size of the municipality where schools are located and is designed to ensure that the sampling process adequately represents the full range of schools over the entire nation. A random selection of schools is drawn on the basis of the number of students in each cell proportionate to the number of students in the universe, creating a cross-section of young people in a set of designated grades (in this study, grades three through 12). The sample design also permits oversampling by a variety of criteria (e.g., location, urbanicity, grade level, school type, or race/ethnicity). This study includes an oversample of schools in areas with a high proportion of African American and Hispanic students.
Once schools were selected as part of the sample, a letter was sent to principals soliciting their participation. Harris Interactive then contacted the principal by telephone to request his or her participation in the survey. A number of steps were included in the consent procedure in order to maximize participation.