The fieldwork of each PSFD survey begins roughly half a month before the Chinese New Year, and finishes by the end of June in the same calendar year. Many important traditional holidays for Chinese society fall within this period, such as the Chinese New Year, and Tomb-Sweeping Day. This period is chosen because this may increase the opportunities to contact or regain connections with the respondents. The preparations for each survey usually begin 1.5 years prior to the fieldwork. At the stage of questionnaire design, the project team scrutinizes the existing questionnaires, and revises the core modules carefully with the premise that intertemporal comparison for the same question is feasible.
The importance of the appended questions, their relationships with the existing core modules, and the length of the questionnaire are the main concerns when it comes to the design of the appended modules. After the preliminary design of the questionnaire is finished, pretests and/or cognitive interviews are conducted if necessary to check whether the newly appended questions and the revised core questions are appropriate. The PSFD team determines whether to use the complete questionnaire or parts of the questionnaire, and the survey mode (e.g., face-to-face interviews or telephone interviews) in pretests. After the cognitive interviews and the pretests, the research team make further revisions based on the results of the cognitive interviews and pretests, until the questionnaires are properly done.
Since 2014, according to the newly established research ethics standards in Taiwan, the PSFD team sends the survey project (including contents of questionnaires and planning for survey implementation, data protection and release) to the IRB on Humanities and Social Science Research, Academia Sinica for review before data collection. No data collection is made before approval from the IRB.
The fieldwork of the pretests and formal surveys is jointly conducted with the Survey Research & Development (SROD) team, Center for Survey Research, Academia Sinica. The main tasks of the SROD are interviewer recruitment and training, and fieldwork monitoring. In the fieldwork of each survey, more than half of the interviewers had participated in previous PSFD surveys. A three-day interviewer training course is scheduled before the fieldwork of each survey. In the training, the PSFD team explains and discusses the contents of the questionnaires as well as the rules for making records, while the SROD is in charge of the other training courses such as the principles of interviewing and ethical requirements in the fieldwork. The other tasks of the SROD include sending the advance letters to the targeted respondents, and issuing official documents to the relevant public institutions (e.g., National Police Agency, Ministry of the Interior, 165 anti-fraud hotline, local police stations, and village offices) for the respondents to verify when necessary. The advance letters designed by the PSFD team contain information on the purpose of the survey, importance of participation, substantial incentive, estimated time of interview, scheduled fieldwork period, channels of survey verification, and contact information of the PSFD staff.
From 1999 to 2016, almost all questionnaires were completed via face-to-face interview. Since the 2018 survey, mixed modes have been adopted, with online self-administered questionnaires as an additional mode of data collection. As to face-to-face interviews, between 1999 and 2011, paper questionnaires were used; from 2012 onwards, PSFD introduced the Computer-Assisted Personal Interviewing (CAPI) system developed by the Center for Survey Research (CSR) to conduct in-person interviews. Since 2018, the Computer-Assisted Web Interviewing (CAWI) system developed by CSR has been used in online data collection of the PSFD surveys.
To ensure the quality of the complete questionnaires, the SROD randomly draws some respondents to re-interview via phone call or email. For the selected respondents who are not reachable by phone or email, a face-to-face re-interview is arranged with a different interviewer. Starting in 2020, face-to-face interviews are recorded with the respondent’s consent. After adopting this new policy, few telephone re-interviews have been needed for the cases completed by face-to-face interview. Respondents who finish an online questionnaire are rechecked by telephone, without exception.
During the fieldwork, the PSFD team roughly checks the collected data. After the fieldwork, the data are cleaned further for possible illogical, inconsistent, or illegal responses. In addition, data from previous waves of survey are used to check for inconsistencies among waves. Metadata (e.g., codebooks) are also produced.
Protocol for Follow-ups
The first-wave questionnaire of the main respondents (Questionnaire RI) contains more questions than the follow-ups. The main respondents who finish the first-wave survey are contacted for follow-up surveys using Questionnaire RR. These follow-up surveys were conducted annually before 2012. Since 2012, these follow-up surveys have been conducted every two years.
For the original main respondents who finish the first-wave questionnaire, the PSFD team attempt to contact them in the follow-up surveys. However, sample attrition might occur due to permanent (e.g., death) or temporary reasons (e.g., inability to contact them, or illness). For temporary non-respondents, the PSFD team attempts to contact each respondent in the subsequent waves of survey. As to the child sample, the PSFD team tries to contact them once they reach age 16. The child respondents who finish their first interview are traced in the follow-up surveys, regardless of whether the original main respondents are still in the sample. However, for the child sample who haven’t been contacted before, the PSFD team does not attempt to contact them if the original main respondents have dropped out of the survey for good.