National Health and Nutrition Examination Survey

NHANES 1999-2000 Questionnaire Data Overview

Respondent Rules

In general, persons 16 years of age and older, and emancipated minors were interviewed directly. A small number of NHANES questionnaire sections have different respondent rules. Please review the questionnaire documentation, questionnaires, and interviewer training manuals carefully prior to analyzing the data.

Proxy Rules

A responsible adult provided information for sample persons under 16 years of age and for individuals who could not self-report. The interviewer training manuals and component-specific notes provided with these files provide additional information about the use of respondents.

Automated Data Collection Methodology

Household Interview and Mobile Examination Center (MEC) Interview data were recorded electronically with the use of computerized questionnaire forms that were programmed using Blaise© software. The Household Interview questionnaires were programmed in Blaise© and installed on laptop computers for use in the field. The MEC interviews were also administered in an automated format, but the interview module formats included non-Blaise© software. After data collection was completed, the interview data files were transmitted electronically to a central survey database system.

NHANES Spanish Language Questionnaires

All NHANES questionnaires were translated into Spanish and were administered in computer-assisted personal interview (CAPI) format along with the English language versions.

Household Interview Component Questionnaires:

  • Screener Questionnaire
  • Family Interview Questionnaire
  • Sample Person Questionnaire

Mobile Examination Center (MEC) Questionnaires:

  • MEC Computer-Assisted Personal Interview (MEC CAPI)
  • MEC Audio-Computer-Assisted Self Interview (MEC A-CASI)

The MEC Interview component is a two-part interview that addresses several health-related topics. The personal interview portion is conducted in a private setting by a trained interviewer. The Audio Computerized Self Administered (A-CASI) portion of the interview was conducted in a private room in the MEC without the interviewer being present. (The interviewers are present for a brief "training" session dealing with the use of the personal computer and interview system software) The A-CASI format permits the respondents to hear questions through earphones and to read questions on a computer screen. Respondents progress through the A-CASI questions at their own speed, as they touch the computer screen to indicate their response. The A-CASI portion of the interview consists of five separate sections, each of which deals with a sensitive health risk behavior topic.

Interviewer Training:

All interviewers completed a comprehensive two-week training program. Many of the interviewers had prior interviewing experience. A large percentage of the household interviewers were bilingual in English and Spanish.
The MEC interviewers also received extensive interviewer training on personal and audio-computer-assisted interview administration. This training focused on the specific interview modules that were administered in the MEC dealing with mental health and smoking, drug use, alcohol, and sexual behavior.

NHANES interviewer training included role-playing exercises and practice interviews, all of which were monitored by NCHS and contractor staff. Detailed information about the interviewer training requirements is found in the NHANES Household Interviewer Training Manual.

Interviewer Quality Control Monitoring

Extensive pretesting was completed prior to implementing the questionnaires in the field. A full "dress rehearsal" pilot test was conducted with the field and examination staff to test all of the survey participant contact and interview systems and procedures.

Several types of quality control monitoring methods were implemented to ensure that high quality data were collected during the survey. Interviewer monitoring was a major responsibility for NCHS and contractor staff. Interviewers were frequently accompanied on interviews and observed to verify that the interview protocol was administered correctly. Interviewers were retrained on survey procedures if necessary.

NCHS encouraged the interviewers to provide constructive comments and feedback based on their field experiences. NCHS used field staff and interviewer feedback to improve the questionnaires and survey materials. Survey staff debriefing and training sessions were conducted annually. NCHS and contractor staff participated in the debriefing meetings. Interviewers were trained on new survey content and protocol changes prior to implementation in the field.

Data collection system consistency checks

The NHANES CAPI systems have built-in edit and range checks for most questions that have open-ended response options. When unusual or unrealistic responses were recorded, the interviewer was alerted immediately and instructed to verify or edit the initial response.

Data Preparation: Reviews, editing, and derived variables

During data preparation, variable frequency counts were checked, questionnaire "skip" patterns were verified, and the reasonableness of responses to the questions were reviewed. If changes were made to the original data, a derived variable was created. The derived variable is essentially the final version of the variable after editing and recoding was performed. The derived variable nomenclature for questionnaire data includes the letter "D" in the third position of the variable name in place of the letter "Q" for the questionnaire item that is customarily used in NHANES data release files. The survey questionnaire codebooks include the names and descriptions of the derived variables.

Check Items and Final Data Release Variables

The codebooks for each survey component include "check item" variables. Check item variables were used internally at NCHS as part of the quality control process to verify that the data collection process was correct. The check item variables are not part of the data release files, however. Analysts would need to re-create these variables to produce check item frequencies