The 1998 Healthy Longevity Survey was the first nationwide survey on the oldest old in China. It was a very demanding task because of the length of the questionnaire, the scattered geographic distribution of the sampled elders and their poor health condition, and the inclement weather in most sampled areas. To guarantee quality, the survey was carefully planned and closely supervised.

Based on the sample frames, the name list of the respondents in each selected area was prepared through various channels like the household registration institutions, village (residential) committees, and the grassroots aging organizations. The age of the elders was verified to determine the samples. A waiting list was prepared to substitute those who might die before the interview.

The quality of the interviewers was decisive for the success of the survey. Interviewers were selected according to such criteria as familiarity with the elders’ work, having high education and relevant survey experience, carefulness and industriousness, and caring for and patient with the elders. The medical staff in the survey team was also carefully selected.

The interviewers were trained both at the national and provincial levels through lecturing, imitated interviews, discussions and examinations. Videotapes based on the experiences gained from the pilot surveys were produced for the training. Survey organizers and supervisors from 22 provinces were first trained, and then they practiced and passed examinations at the national training workshop. They organized training workshops in each province. The videotapes for training interviewers and detailed written instructions were distributed. Training at all 22 provinces was thus standardized, which was crucial for ensuring uniform interpretations of the questionnaire and the good quality of the survey. About 300 interviewers were trained on the survey objectives, survey design, questionnaire items, interview skills, and other requirements. 120 survey teams were organized nationwide, with one interviewer, one recorder and one medical staff in each team.

The survey was carried out by the provincial aging associations, with strong support and help from local governments, which combined the survey with caring for the elders. 120 survey bags were distributed to survey teams, with questionnaire, survey instructions, physical check equipment, camera and flashlight inside. Special transportation was arranged for those elders living in remote areas. 1 or 2 guides were prepared for each team for household interviews and interpreters were provided when needed.

The field survey started in June 1998 and accomplished by December 1998. 9073 valid individual questionnaires were completed and 4116 DNA samples were collected on an entirely voluntary basis. The dedication of interviewers to the survey was exemplary. They overcame various difficulties and tried their best to verify each item and match the centenarians to the highest standard.

The survey was warmly welcomed and actively coordinated by the interviewees and their families. Officers from county or township aging associations accompanied the survey teams to visit the elders, showing their concern for the elders, giving them gifts and other material benefits, learning their difficulties in daily life, and concretizing beneficial policies for the elders.

To guarantee survey quality, detailed error check and quality control mechanisms were made beforehand. For each elder interviewed, the interviewers were required to “self check the questionnaire, mutually check it, check it by sampling and check the internal consistency”. Problems were found out immediately on the site and corrected. Every survey team was required to go to each household in the sample and avoid omitting any household, anyone in the household, or any item in the questionnaire. The completed questionnaires were checked by the survey supervisors, by the provincial survey organizers, and by the project members. The project members also went to the field for supervision. All these efforts ensured the high quality of this survey.