Overview and Methodology of the Adolescent Behaviors and…

ABES Methodology

Overview

ABES was a one-time, online survey conducted during January–June 2021 to assess student behaviors and experiences during the COVID-19 pandemic. ABES surveyed high school students in grades 9–12 attending U.S. public and private schools. Classes were randomly selected to participate within a nationally representative sample of schools. Because of the different instructional models used across the nation during the pandemic (i.e., in-person only, virtual only, and hybrid), ABES was designed as a self-administered, anonymous survey that was administered online. The online administration allowed each school and teacher the flexibility to decide whether students completed the survey during instructional time or on their own time. In addition to student-level data, school-level data (e.g., instructional mode and school-provided equipment) also were collected.

School Characteristics

The COVID-19 pandemic resulted in modifications of the learning environment, and many students faced several challenges, including lack of adequate resources at home (7). Therefore, collecting information on school characteristics during the fielding of ABES was important. School staff reported the school’s instructional model (in-person only, virtual only, or hybrid) at the time of recruitment; hybrid schools had a portion of a school’s student body receiving in-person instruction while others received virtual instruction. School staff members also reported whether the school had reduced class sizes and whether they provided laptops or Chromebooks, tablet computers, and Wi-Fi hotspots to students for home use.

Student Online Questionnaire and Administration

Students completed ABES questionnaires in English or Spanish using a secure URL on any Internet-connected device. They were required to watch a 2-minute video (with an option of an English version, with English audio, or a Spanish version, with English audio and Spanish subtitles) and read brief instructions before starting the questionnaire. On average, students completed the 110-item questionnaire in 30 minutes. Although students were encouraged to finish the survey in one sitting, they were able to save their answers, stop the survey, and resume the survey as needed. Cumulative timing data were captured for the entire time a student was logged in, thus a meaningful range of completion time is not available. The survey included 97 questions from the 2021 national YRBS questionnaire. Six of these questions were modified to allow students who were attending school only virtually to indicate that a question asking about a behavior on school property did not apply. The questionnaire also included 12 new questions (not included on the YRBS questionnaire) assessing COVID-19–related behaviors and experiences and one new question on perceived racism (Box). The questionnaire was designed with no skip patterns; however, students could skip questions they did not want to answer. (Questionnaires in English and Spanish are available at https://www.cdc.gov/healthyyouth/data/abes.htm.)

In participating schools, after parental permission was granted and a student agreed to participate, teachers of selected classes provided students with instructions for accessing the survey and a randomly generated login code that allowed for completely anonymous participation. Among eligible students, 2.6% had a parent who refused to allow participation, and an additional 4.3% did not return a permission form when active consent was required. To help maintain privacy, only one question appeared on the screen at a time. Images were added for questions related to use of tobacco products, substances, and contraceptives to improve respondents’ recognition and understanding of the questions.

Sampling

ABES used the same sampling methods as the national YRBS, except that a larger sample was drawn in anticipation of lower response rates (6). The ABES sampling frame included public and private schools with grades 9–12 in all 50 U.S. states and the District of Columbia. Combined data obtained from MDR (formerly Market Data Retrieval) and the National Center for Education Statistics (NCES) were used to create the sampling frame. Public schools were identified from the Common Core of Data (https://nces.ed.gov/ccdexternal icon), and private schools were identified from the Private School Survey (https://nces.ed.gov/surveys/pssexternal icon), both of which are NCES databases. The sampling frame excluded alternative, special education, U.S. Department of Defense, Bureau of Indian Education, and vocational schools that serve students who are also enrolled in another public school. Home-schooled students and students who dropped out of high school were not eligible for ABES if they were not enrolled in a school.

ABES used a stratified, three-stage cluster sampling approach to obtain a nationally representative sample of students. The first sampling stage consisted of primary sampling units (PSUs) encompassing a county, portion of a county, or a group of counties. A total of 81 PSUs were selected across 16 primary strata by urban and nonurban location and percentage of non-Hispanic Black and Hispanic or Latino (Hispanic) students in the PSU. The second sampling stage consisted of 335 secondary sampling units (SSUs), defined as physical schools with grades 9–12 or linked schools combined to provide the four grades. At the first and second stages, the sample was selected with probability proportional to size. Schools with ≥28 students per grade were defined as large, and those with <28 students were defined as small. The third sampling stage selected one or two classes within each grade of an SSU. All students in a selected class were eligible for ABES, unless they were unable to complete the questionnaire independently.

Data Collection Procedures

The ABES study protocol was reviewed and approved by institutional review boards at CDC and ICF International, CDC’s survey contractor.* The survey was administered during January–June 2021; student participation was voluntary and anonymous. ABES was designed to be a self-administered questionnaire completed online during or outside of instructional time on any Internet-connected device. Most participating students (91%) completed ABES using a laptop or desktop computer; 9% of students used a phone or tablet computer.

Data Processing and Response Rates

ABES data were cleaned and edited for inconsistencies. Values that were not plausible or logical were excluded from analysis (e.g., a student who answered that he had never smoked cigarettes but also answered that he had smoked cigarettes during the past 30 days). Before data editing, students in 128 schools submitted 7,998 ABES questionnaires (7,953 in English and 45 in Spanish). Among these 7,998 records, 293 were excluded because <20 questions had been answered, resulting in 7,705 questionnaires with valid data. The school response rate was 38%, the student response rate was 48%, and the overall response rate ([Student response rate] × [School response rate]) was 18%.

Race and ethnicity were ascertained using methods from the 2019 national YRBS (6); however, in this MMWR Supplement, the “other” race and ethnicity category was disaggregated to report the following racial categories (all non-Hispanic): American Indian or Alaska Native, Asian, Native Hawaiian or other Pacific Islander, and multiracial. The response options for sexual identity differed from those in the 2019 national YRBS (Table 1).

Weighting

ABES data were weighted based on student sex and grade to account for school and student nonresponse and the oversampling of non-Hispanic Black students and Hispanic students. Weights were applied to all records and scaled so that weighted counts equaled the total sample size, and weighted student proportions in each grade matched national proportions. As a result, ABES weighted estimates are nationally representative of all students in grades 9–12 attending public and private schools in the United States.

Analytic Methods

To account for the complex sampling design and weighting, ABES data were analyzed using SAS (version 9.4; SAS Institute) and SUDAAN (version 11.0.1 or 11.0.3; RTI International). The methods section of each report in this MMWR Supplement includes details on the analytical methods used in each analysis. For this overview report, a descriptive analysis of ABES student- and school-level data was conducted.

Data Availability and Dissemination

ABES data and documentation describing the data are available to the public (https://www.cdc.gov/healthyyouth/data/abes.htm). Student survey and school characteristics data have been merged into one data set and are available in Access and ASCII formats. SAS and SPSS programs are provided for converting the ASCII data into SAS and SPSS data sets. No approval is needed to use the ABES data set. Data requests and other ABES-related questions can be sent to CDC by using the YRBS data request form (https://www.cdc.gov/healthyyouth/data/yrbs/contact.htm).